Brain imaging analysis method, brain imaging analysis device, and brain imaging analysis system

Abstract

To provide a medical image analysis method.SOLUTION: A medical image analysis method includes: acquiring a correction parameter calculated based on a correlation between a first morphological value acquired from a first medical image acquired under a first scan condition and associated with a target element, and a second morphological value acquired from a second medical image acquired under a second scan condition and associated with the target element; acquiring a target medical image acquired under the second scan condition; executing segmentation on the target medical image in a plurality of regions corresponding to a plurality of elements including the target element to acquire a target region associated with the target element from the target medical image; acquiring a target morphological value associated with the target element based on voxel data corresponding to the target region; acquiring a corrected morphological value based on the target morphological value and the correction parameter; and outputting a morphological index based on the corrected morphological value.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present application relates to a medical video analysis method, a medical video analysis device, and a medical video analysis system for analyzing medical videos.

Background Art

[0002] With the improvement of image segmentation technology, it has become possible to segment medical videos and calculate diagnostic assistance indicators related to diseases, etc., and the field of recent medical video analysis has attracted attention.

[0003] In particular, medical video analysis technology is widely used to provide diagnostic assistance indicators for dementia, and it is essential to more accurately calculate the morphological characteristics of specific regions from medical videos in order to provide objective dementia diagnostic assistance information.

[0004] However, conventional medical video analysis technology has various limitations due to errors in the medical video itself, such as containing artifacts that are not suitable for image segmentation in the medical video, and there is a limitation that the diagnostic assistance information for the same object may vary depending on the scan conditions under which the medical video is acquired. In addition, conventional medical video analysis technology has a problem that it cannot provide completely "personalized" diagnostic assistance information for the object by using a method of performing image segmentation after standardizing the brain contained in the medical video with respect to the standard brain model, and in a situation where high accuracy is required, it is in a situation with limitations.

[0005] Therefore, there is a need to develop an image analysis method and device that can be provided to the user in the case of diagnostic information that can be determined even when the image contains noise. There is a need to develop an image analysis method and device for obtaining diagnostic assistance information in consideration of the scan conditions under which the medical video was taken and obtaining personalized diagnostic assistance information for the object.

Summary of the Invention

[0006] One objective of the present invention is to provide a medical video analysis method, a medical video analysis device, and a medical video analysis system that provide information related to medical images.

[0007] The problems that this invention aims to solve are not limited to those described above, and any problems not mentioned should be clearly understood by those skilled in the art from this specification and the drawings. [Means for solving the problem]

[0008] The medical image analysis method disclosed in this application may include: obtaining a correction parameter calculated based on the correlation between a first morphological value obtained from a first medical image obtained under first scan conditions and associated with a target element, and a second morphological value obtained from a second medical image obtained under second scan conditions and associated with the target element; obtaining a target medical image obtained under the second scan conditions; obtaining a target region associated with the target element from the target medical image by performing segmentation of the target medical image into multiple regions corresponding to multiple elements including the target element; obtaining a target morphological value associated with the target element based on voxel data corresponding to the target region; obtaining a morphological value corrected based on the target morphological value and the correction parameter; and outputting a morphological index based on the corrected morphological value.

[0009] The medical image analysis apparatus disclosed in this application comprises an image acquisition unit that acquires a target medical image, and a controller that provides medical image analysis information based on the target medical image, wherein the controller may be configured to acquire a correction parameter calculated based on the correlation between a first morphological value acquired from a first medical image acquired under first scan conditions and associated with a target element and a second morphological value acquired from a second medical image acquired under second scan conditions and associated with the target element, acquire a target medical image acquired under the second scan conditions, perform segmentation of the target medical image into a plurality of regions corresponding to a plurality of elements including the target element to acquire a target region associated with the target element from the target medical image, acquire a target morphological value associated with the target element based on voxel data corresponding to the target region, acquire a morphological value corrected based on the target morphological value and the correction parameter, and output a morphological index based on the corrected morphological value.

[0010] The medical image analysis method disclosed in this application includes acquiring a target medical image; acquiring target scan conditions associated with the target medical image; acquiring target morphological values ​​associated with the target element based on voxel data of a target region corresponding to a target element contained in the target medical image; determining a target correction parameter from one or more correction parameters based on the target scan conditions; and outputting a morphological index based on the determined target correction parameter and the target morphological values, wherein determining the target correction parameter may include, if the target scan conditions correspond to a first scan condition, the target correction parameter being determined by a first correction parameter for correcting a first morphological value associated with the target element acquired under the first scan conditions; and if the target scan conditions correspond to a second scan condition, the target correction parameter being determined by a second correction parameter for correcting a second morphological value associated with the target element acquired under second scan conditions different from the first scan conditions.

[0011] The medical image analysis apparatus disclosed in this application comprises an image acquisition unit that acquires a target medical image, and a controller that provides medical image analysis information based on the target medical image, wherein the controller is configured to acquire target scan conditions related to the target medical image, acquire target morphological values ​​related to the target element based on voxel data of a target region corresponding to a target element included in the target medical image, determine a target correction parameter from one or more correction parameters based on the target scan conditions, and output a morphological index based on the determined target correction parameter and the target morphological value, wherein the target correction parameter may be determined by a first correction parameter for correcting a first morphological value related to the target element acquired under the first scan conditions when the target scan conditions correspond to the first scan conditions, and by a second correction parameter for correcting a second morphological value related to the target element acquired under second scan conditions different from the first scan conditions when the target scan conditions correspond to the second scan conditions.

[0012] The means for solving the problems of the present invention are not limited to the solutions described above, and solutions not mentioned should be clearly understood by those skilled in the art from this specification and the accompanying drawings. [Effects of the Invention]

[0013] According to the embodiment of this application, a quality analysis can be performed on the target medical image, and information regarding the analysis results can be provided to the user to increase the reliability of the medical image analysis results.

[0014] According to the embodiment of this application, morphological indicators can be calculated more accurately by calculating morphological indicators based on a medical image of the target body and appropriately applying correction parameters considering the scanning conditions under which the medical image is acquired and the position of the target elements.

[0015] According to the embodiment of this application, it is possible to selectively provide the user with the necessary indicator information from among the various indicator information obtained through the analysis of the target medical image, thereby enhancing ease of use.

[0016] The effects of the present invention are not limited to those described above, and any effects not mentioned herein should be clearly understood by those skilled in the art from this specification and the accompanying drawings. [Brief explanation of the drawing]

[0017] [Figure 1] This is a schematic diagram relating to a medical image analysis system according to one embodiment of this application. [Figure 2] This is a schematic diagram of an image analysis device 2000, a learning device 2200, a correction parameter acquisition device 2400, and an output device 2600 for image analysis according to one embodiment of the present application. [Figure 3] This is a block diagram relating to an image analysis device according to one embodiment of this application. [Figure 4] This is a block diagram of an output device according to one embodiment of the present application. [Figure 5] This figure shows an example of the image alignment operation of the image analysis device 2000. [Figure 6] This figure shows an example of the image alignment operation of the image analysis device 2000. [Figure 7] This figure shows the process flow for image segmentation according to one embodiment of the present application. [Figure 8] This is a step-by-step diagram illustrating a method for learning a neural network model using a learning device 2200 according to one embodiment of this application. [Figure 9] This is an illustrative structural diagram of an image dataset according to one embodiment of this application. [Figure 10] This is an example of an artificial neural network model that can be used by the learning device 2200 according to an embodiment of the present application. [Figure 11] This is another example of an artificial neural network model that can be used by the learning device 2200 according to an embodiment of the present application. [Figure 12] This is a sequence diagram for an image segmentation method using the neural network model of the image analysis device 2000 according to an embodiment of the present application. [Figure 13] This is an exemplary structural diagram of a target image according to an embodiment of the present application. [Figure 14] This is an exemplary diagram of an image obtained based on the result value acquired by the segmentation process of the image analysis device 2000 according to an embodiment of the present application. [Figure 15] This is a diagram showing the flow of a process for image segmentation according to an embodiment of the present application. [Figure 16] This is an exemplary structural diagram of an image dataset according to an embodiment of the present application. [Figure 17] This is a diagram showing the flow of a process for image segmentation according to an embodiment of the present application. [Figure 18] This is a diagram for explaining an image quality determination process according to an embodiment. [Figure 19] This is a diagram for explaining an image analysis device that executes an image quality determination process. [Figure 20] This is a diagram for explaining an image quality determination module. [Figure 21] This is a diagram for explaining a first image quality determination process. [Figure 22] This is a diagram for explaining a first image quality determination unit. [Figure 23] This is a diagram for explaining a second image quality determination process. [Figure 24] This is a diagram for exemplarily explaining that the second image quality determination process uses an artificial neural network model to determine the position where artifacts occur in a medical image. [Figure 25] This is a diagram illustrating a neural network model for acquiring artifact information according to one embodiment. [Figure 26] This is a diagram illustrating the third image quality assessment process. [Figure 27] This is a diagram illustrating one embodiment of the third image quality judgment process. [Figure 28] This is a diagram illustrating the fourth image quality assessment process. [Figure 29] This is a diagram illustrating an image where an artifact has occurred. [Figure 30] This figure illustrates how, in the fourth image quality assessment process according to other embodiments, it is determined whether an artifact occurred overlapping with the region of interest. [Figure 31] This figure illustrates an image quality-related information output module according to one embodiment. [Figure 32] This is a diagram illustrating the first error information screen. [Figure 33-34] This is a diagram illustrating the second error information screen. [Figure 35-36] This is a diagram illustrating the third error information screen. [Figure 37-38] This is a diagram illustrating the fourth error information screen. [Figure 39] This diagram illustrates the selection window output by the image quality-related information output module 3900. [Figure 40] This diagram illustrates the selection window output by the image quality-related information output module 3900. [Figure 41-43] This is a diagram illustrating the first to third quality assessment processes. [Figure 44-45] This is a diagram illustrating the fourth quality assessment process. [Figure 46-47] This is a diagram illustrating the fifth quality assessment process. [Figure 48]This flowchart illustrates one operation of an image analysis method implemented by the image analysis apparatus 2000 according to one embodiment of this application. [Figure 49] This is an illustrative diagram of an image alignment method realized by an image analysis device 2000 according to one embodiment of this application. [Figure 50] This is a flowchart of an image analysis method according to one embodiment of this application. [Figure 51] This figure shows an example of a method for modifying the first boundary of the first internal region of the skull according to one embodiment of the present invention. [Figure 52] This figure shows an example of a second internal region obtained by a method for modifying the first boundary of a first internal region of a skull according to one embodiment of this application. [Figure 53] This figure shows another example of a method for modifying the first boundary of a first internal region of the skull according to one embodiment of the present application. [Figure 54-57] This is an illustrative figure showing morphological indicators and related information output by one embodiment of this application. [Figure 58] This flowchart illustrates one operation of a method for correcting morphological values ​​realized by an image analysis device 2000 according to one embodiment of this application. [Figure 59] This is a flowchart of an image analysis method according to one embodiment of this application. [Figure 60] This flowchart shows one operation of a method for acquiring correction parameters realized by the correction parameter acquisition device 2400 according to one embodiment of this application. [Figure 61] This is an illustrative diagram illustrating the information contained in multiple image datasets according to one embodiment of this application. [Figure 62] This graph shows an example of a correlation analysis for obtaining correction parameters using a correction parameter acquisition device 2400 according to one embodiment of this application. [Figure 63] This graph shows another example of correlation analysis for obtaining correction parameters using the correction parameter acquisition device 2400 according to one embodiment of this application. [Figure 64]This is an illustrative diagram illustrating the information contained in multiple image datasets according to one embodiment of this application. [Figure 65] This is an illustrative diagram of a correlation analysis between morphological values ​​corresponding to the distribution of target elements within the skull, according to one embodiment of the present application. [Figure 66] This is a flowchart of an image analysis method according to one embodiment of this application. [Figure 67] This is a schematic diagram illustrating a user interface according to one embodiment of the present application. [Figure 68] This is a diagram illustrating a medical information output process according to one embodiment. [Figure 69] This is a diagram illustrating an image analysis apparatus according to one embodiment. [Figure 70] This is a diagram illustrating the medical data acquisition module. [Figure 71] This is a diagram illustrating the medical information acquisition module. [Figure 72] This is a diagram illustrating the diagnostic support information acquisition module. [Figure 73] This is a diagram illustrating the diagnostic support information output module. [Figure 74] This diagram illustrates relevant indicators for several target diseases. [Figure 75] This diagram illustrates a medical information output screen according to one embodiment. [Figure 76] This diagram illustrates the medical information output process in another embodiment. [Figure 77] This figure illustrates an image analysis apparatus according to another embodiment. [Figure 78] This is a diagram illustrating the sorting information acquisition module. [Figure 79] This is a diagram illustrating the sorting information output module. [Figure 80-81] This is a diagram illustrating the sorting information output screen. [Modes for carrying out the invention]

[0018] A medical image analysis method according to one embodiment of the present application includes the steps of: acquiring a target medical image containing at least one artifact; acquiring artifact information associated with the artifact based on the target medical image, wherein the artifact information includes an artifact region indicating the region in which the artifact is distributed; acquiring a plurality of regions based on the target medical image using a first neural network model, wherein the first neural network model is trained to acquire the plurality of regions based on the medical image, wherein the plurality of regions include a first region corresponding to a first element, a second region corresponding to a second element, and a third region corresponding to a third element; acquiring the degree of overlap between the artifact region and the region of interest, wherein the region of interest includes at least a portion of the first region; a first quality determination step of performing a quality determination of the target medical image based on the degree of overlap; and outputting quality-related information of the target medical image based on the first quality determination.

[0019] The first element is a skull, the first region includes a region corresponding to the skull, and the region of interest may be the skull region or an internal region of the skull.

[0020] The second element is an element related to the diagnosis of the target disease, and the area of ​​interest may include at least a part of the second area.

[0021] The artifact region is obtained using a second neural network model trained to determine whether or not an artifact exists in the medical image, and the artifact region may be a region on the feature map obtained from the second neural network model and associated with the target artifact present in the target medical image, where the association with the target artifact is greater than or equal to a threshold value.

[0022] The artifact region includes a first region and a second region located within the first region, wherein the first region is a region on the feature map whose relationship with the target artifact is greater than or equal to a first criterion value, and the second region may be a region on the feature map whose relationship with the target artifact is greater than or equal to a second criterion value which is greater than the first criterion value.

[0023] The artifact region can be obtained using a second neural network model that has been trained to segment the region corresponding to the artifact contained in the medical image.

[0024] The step of determining the degree of overlap may include obtaining the outline of the artifact region and the outline of the region of interest on the target medical image, and determining whether the artifact and the region of interest overlap based on whether the outline of the artifact region and the outline of the region of interest overlap on the medical image.

[0025] The first quality determination step may include determining that the quality of the target medical image is normal if there are fewer than two intersections where the outline of the artifact region and the outline of the region of interest overlap on the target medical image, and determining that the quality of the target medical image is abnormal if there are two or more such intersections.

[0026] The step of determining the degree of overlap may include the steps of obtaining the number of common pixels that are fully included in the artifact region and the region of interest on the target medical image, and determining the degree of overlap between the artifact and the region of interest using the number of common pixels.

[0027] The first quality determination step may include determining that the quality of the target medical image is abnormal if the number of common pixels exceeds a predetermined standard value, and determining that the quality of the target medical image is normal if the number of common pixels is less than or equal to a predetermined standard value.

[0028] The process further includes a second quality assessment step of performing a quality assessment of the target medical image based on the third domain, the second quality assessment step may include a step of obtaining morphological indicators based on the third domain and a step of determining whether the morphological indicators meet quantitative criteria.

[0029] The morphological indicators may be obtained based on the first and third regions.

[0030] The morphological index may be obtained based on the ratio of the volume of the first region to the volume of the third region.

[0031] The third region may correspond to the element located at the bottom of the target medical image.

[0032] The third area described above may correspond to elements related to the diagnosis of the target disease.

[0033] The step of outputting quality-related information of the medical image includes displaying error information generated based on the first quality judgment and a selection window associated with the error information, wherein the error information may include the artifact information and information about the plurality of regions.

[0034] The selection window includes a first object, and the medical image analysis method includes proceeding with subsequent actions according to the user's selection of the first object, the subsequent actions may include any of the following: the target medical image analysis action, the action to correct the target medical image, or the action to re-capture the target medical image.

[0035] A medical image analysis device according to another embodiment of the present application includes an acquisition unit for acquiring a medical image, a processing unit for acquiring image quality information based on the medical image, and an output unit for outputting the image quality information, wherein the processing unit acquires a target medical image including artifacts via the acquisition unit, acquires artifact information associated with the artifacts based on the target medical image, the artifact information includes an artifact region indicating the region in which the artifacts are distributed, and uses a first neural network model to acquire a plurality of regions corresponding to one or more anatomically divided elements based on the target medical image, the first neural network model is trained to acquire the plurality of regions based on the medical image, the plurality of regions include a first region corresponding to a first element, a second region corresponding to a second element, and a third region corresponding to a third element, the degree of overlap between the artifact region and the region of interest is acquired, the region of interest includes at least a portion of the first region, a first quality determination of the target medical image is performed based on the degree of overlap, and the output unit can output quality-related information of the target medical image based on the first quality determination.

[0036] The first element is a skull, the first region includes a region corresponding to the skull, and the region of interest may be the skull region or an internal region of the skull.

[0037] The second element is an element related to the diagnosis of the target disease, and the area of ​​interest may include at least a portion of the second area.

[0038] The artifact region is obtained using a second neural network model trained to determine whether or not an artifact exists in the medical image, and the artifact region may be a region on the feature map where the association with the target artifact is greater than or equal to a threshold value, based on a feature map obtained from the second neural network model and associated with the target artifact present in the target medical image.

[0039] The artifact region includes a first region and a second region located within the first region, wherein the first region is a region on the feature map whose relationship with the target artifact is greater than or equal to a first criterion value, and the second region may be a region on the feature map whose relationship with the target artifact is greater than or equal to a second criterion value which is greater than the first criterion value.

[0040] The artifact region may be obtained using a second neural network model that has been trained to segment the region corresponding to the artifact contained in the medical image.

[0041] The processing unit can obtain the outline of the artifact region and the outline of the region of interest on the target medical image, and determine whether the artifact and the region of interest overlap based on whether the outline of the artifact region and the outline of the region of interest overlap on the medical image.

[0042] The processing unit can determine that the quality of the target medical image is normal if there are fewer than two intersections where the outline of the artifact region and the outline of the region of interest overlap on the target medical image, and that the quality of the target medical image is abnormal if there are two or more such intersections.

[0043] The processing unit can obtain the number of common pixels that are fully included in the artifact region and the region of interest on the target medical image, and use the number of common pixels to determine the degree of overlap between the artifact and the region of interest.

[0044] The processing unit can determine that the quality of the target medical image is abnormal if the number of common pixels exceeds a predetermined standard value, and that the quality of the target medical image is normal if the number of common pixels is less than or equal to the predetermined standard value.

[0045] The processing unit may acquire morphological indicators based on the third region and perform a second quality judgment based on whether the morphological indicators meet quantitative criteria.

[0046] The morphological indicators may be obtained based on the first and third regions.

[0047] The morphological index may be obtained based on the ratio of the volume of the first region to the volume of the third region.

[0048] The third region may correspond to the element located at the bottom of the target medical image.

[0049] The third area described above may correspond to elements related to the diagnosis of the target disease.

[0050] The brain imaging analysis method disclosed in this application may include: acquiring a brain image including voxel data; performing segmentation of the brain image into a plurality of regions including a skull region and a reference region to acquire a first boundary defining a first internal region which is inside the skull region and the reference region; acquiring a second internal region having a second boundary by modifying at least a portion of the first boundary related to the first internal region based on the reference region, wherein a portion of the second internal region is included in the first internal region and includes the target element; acquiring a first volume value associated with the target element based on voxel data corresponding to the target element; acquiring a second volume value associated with the second internal region based on voxel data included in the second boundary; and calculating a volume index of the target element based on the first volume value and the second volume value.

[0051] The brain imaging analysis method disclosed in this application may further include aligning the brain images in order to obtain the second internal region.

[0052] According to the brain image analysis method disclosed in this application, aligning the brain images may include obtaining a first feature region and a second feature region from the brain images, calculating a first feature point from the first feature region and a second feature point from the second feature region, obtaining the shooting direction of the brain images based on the first and second feature points, and aligning the brain images so that the shooting direction is parallel to a reference direction.

[0053] According to the brain image analysis method disclosed in this application, the first feature region is the region corresponding to the anterior commissure, the second feature region is the region corresponding to the posterior commissure, the first feature point may be included in the boundary defining the region corresponding to the anterior commissure, and the second feature point may be included in the boundary defining the region corresponding to the posterior commissure.

[0054] According to the brain image analysis method disclosed in this application, aligning the brain images may include obtaining data related to the orientation of the brain images contained in the brain images, obtaining the shooting direction of the brain images based on the data related to the orientation of the brain images, and aligning the brain images such that the shooting direction corresponds to a reference direction.

[0055] According to the brain imaging analysis method disclosed in this application, obtaining the second internal region may include obtaining a reference plane adjacent to the reference region from the brain imaging, and obtaining the second internal region having the second boundary by modifying a portion of the first boundary of the first internal region based on the reference plane.

[0056] According to the brain imaging analysis method disclosed in this application, the reference plane is a plane parallel to the transverse plane of the brain image, and obtaining the second internal region may include modifying the portion of the first boundary located below the reference plane based on the first internal region located above the reference plane, and obtaining the second internal region having the second boundary based on the modified first boundary.

[0057] According to the brain imaging analysis method disclosed in this application, the reference region is a region corresponding to the cerebellum, and the reference plane can be adjacent to the inferior edge of the boundary defining the corresponding region of the cerebellum.

[0058] According to the brain imaging analysis method disclosed in this application, the reference region is a region corresponding to the cerebellum, and the reference plane can be adjacent to the inferior edge of the boundary defining the corresponding region of the cerebellum.

[0059] According to the brain imaging analysis method disclosed in this application, obtaining the second internal region may include obtaining a first feature region and a second feature region from the brain imaging; calculating a first feature point from the first feature region and a second feature point from the second feature region; obtaining a reference direction connecting the first and second feature points based on the first and second feature points; obtaining a reference plane parallel to the reference direction; and obtaining the second internal region having the second boundary by modifying a portion of the first boundary of the first internal region based on the reference plane.

[0060] According to the brain imaging analysis method disclosed in this application, the reference region is a region corresponding to the neck bones, and the reference plane may be adjacent to the region corresponding to the neck bones and perpendicular to the sagittal plane.

[0061] According to the brain imaging analysis method disclosed in this application, acquiring the second internal region may include modifying the portion of the first boundary located below the reference plane based on the first internal region located above the reference plane, and acquiring the second internal region having the second boundary based on the already modified first boundary.

[0062] According to the brain imaging analysis method disclosed in this application, the volume index may be defined as the ratio of the first volume value to the second volume value.

[0063] According to the brain imaging analysis method disclosed in this application, calculating the volume index of the target element may further include obtaining a first correction parameter for correcting the first volume value, wherein the first correction parameter is obtained based on scan conditions under which the brain image is acquired or the position of the target element, and obtaining a first volume correction value, which includes calculating the first volume correction value based on the first correction parameter and the first volume value; obtaining a second correction parameter for correcting the second volume value, wherein the second correction parameter is obtained based on scan conditions under which the brain image is acquired or the position of the target element, and obtaining a second volume correction value, which includes calculating the second volume correction value based on the second correction parameter and the second volume value; and calculating the volume index based on the first volume correction value and the second volume correction value.

[0064] According to the brain imaging analysis method disclosed in this application, the scanning conditions under which the brain images are acquired can be associated with at least one of the following: the resolution of the brain images acquired by the brain imaging device, the magnetic field strength associated with the resolution of the brain images acquired by the brain imaging device, the manufacturer of the brain imaging device, and setting parameters associated with the form of the magnetic field generated by the brain imaging device.

[0065] The brain imaging analysis device disclosed in this application comprises an image acquisition unit that acquires brain images and a controller that provides brain imaging analysis information based on the brain images, wherein the controller may be configured to acquire brain images including voxel data, perform segmentation of the brain images into a plurality of regions including a skull region and a reference region to acquire a first boundary defining a first internal region which is inside the skull region and the reference region, acquire a second internal region having a second boundary by modifying at least a portion of the first boundary related to the first internal region based on the reference region, a portion of the second internal region which is included in the first internal region and includes the target element, acquire a first volume value related to the target element based on voxel data corresponding to the target element, acquire a second volume value related to the second internal region based on voxel data included in the second boundary, and calculate a volume index of the target element based on the first volume value and the second volume value.

[0066] According to the brain imaging analysis device disclosed in this application, the controller may be configured to align the brain images in order to acquire the second internal region.

[0067] According to the brain imaging analysis device disclosed in this application, the controller may be configured to acquire a first feature region and a second feature region from the brain imaging, calculate a first feature point from the first feature region, calculate a second feature point from the second feature region, acquire the imaging direction of the brain imaging based on the first feature point and the second feature point, and align the brain imaging so that the imaging direction is parallel to the reference direction.

[0068] According to the brain imaging analysis device disclosed in this application, the first feature region is the region corresponding to the anterior commissure, the second feature region is the region corresponding to the posterior commissure, the first feature point may be included in the boundary defining the region corresponding to the anterior commissure, and the second feature point may be included in the boundary defining the region corresponding to the posterior commissure.

[0069] According to the brain imaging analysis device disclosed in this application, the controller may be configured to acquire data related to the orientation of the brain images included in the brain images, acquire the shooting direction of the brain images based on the data related to the orientation of the brain images, and align the brain images so that the shooting direction corresponds to a reference direction.

[0070] According to the brain imaging analysis device disclosed in this application, the controller may be configured to acquire the second internal region by acquiring a reference plane adjacent to the reference region from the brain imaging, and modifying a portion of the first boundary of the first internal region based on the reference plane to acquire the second internal region having the second boundary.

[0071] According to the brain imaging analysis device disclosed in this application, the reference plane is a plane parallel to the transverse plane of the brain image, and the controller may be configured to acquire the second internal region by modifying the portion of the first boundary located below the reference plane based on the first internal region located above the reference plane, and acquiring the second internal region having the second boundary based on the modified first boundary.

[0072] According to the brain imaging analysis device disclosed in this application, the reference region is a region corresponding to the cerebellum, and the reference plane can be adjacent to the inferior edge of the boundary defining the corresponding region of the cerebellum.

[0073] According to the brain imaging analysis device disclosed in this application, the controller may be configured to acquire a second internal region by acquiring a first feature region and a second feature region from the brain imaging, calculating a first feature point from the first feature region, calculating a second feature point from the second feature region, acquiring a reference direction connecting the first feature point and the second feature point based on the first feature point and the second feature point, acquiring a reference plane parallel to the reference direction, and acquiring the second internal region having the second boundary by modifying a part of the first boundary of the first internal region based on the reference plane.

[0074] According to the brain imaging analysis device disclosed in this application, the reference region is a region corresponding to the neck bones, and the reference plane may be adjacent to the region corresponding to the neck bones and perpendicular to the sagittal plane.

[0075] According to the brain imaging analysis device disclosed in this application, the controller may be configured to modify the portion of the first boundary located below the reference plane based on the first internal region located above the reference plane, and to acquire the second internal region having the second boundary based on the modified first boundary.

[0076] According to the brain imaging analysis device disclosed in this application, the controller may be configured to define the volume index as the ratio of the first volume value to the second volume value.

[0077] According to the brain imaging analysis device disclosed in this application, the controller may be configured to calculate the volume index of the target element by acquiring a first correction parameter for correcting the first volume value, the first correction parameter being acquired based on the scan conditions under which the brain image is acquired or the position of the target element, calculating the first volume correction value based on the first correction parameter and the first volume value, acquiring a second correction parameter for correcting the second volume value, the second correction parameter being acquired based on the scan conditions under which the brain image is acquired or the position of the target element, calculating the second volume correction value based on the second correction parameter and the second volume value, and calculating the volume index based on the first volume correction value and the second volume correction value.

[0078] According to the brain imaging analysis device disclosed in this application, the scan conditions under which the brain image is acquired can be associated with at least one of the following: the resolution of the brain image acquired by the brain imaging device and the magnetic field strength associated with it; the manufacturer of the brain imaging device; and setting parameters associated with the form of the magnetic field generated by the brain imaging device.

[0079] The medical image analysis method disclosed in this application may include: obtaining a correction parameter calculated based on the correlation between a first morphological value obtained from a first medical image obtained under first scan conditions and associated with a target element, and a second morphological value obtained from a second medical image obtained under second scan conditions and associated with the target element; obtaining a target medical image obtained under the second scan conditions; obtaining a target region associated with the target element from the target medical image by performing segmentation of the target medical image into a plurality of regions corresponding to a plurality of elements including the target element; obtaining a target morphological value associated with the target element based on voxel data corresponding to the target region; obtaining a morphological value corrected based on the target morphological value and the correction parameter; and outputting a morphological index based on the corrected morphological value.

[0080] According to the medical image analysis method disclosed in this application, the correction parameter includes a parameter for calculating the first morphological value based on the second morphological value, and obtaining the corrected morphological value may include obtaining the corrected morphological value, which is a morphological estimate of the target element under the first scan conditions, based on the target morphological value and the correction parameter.

[0081] According to the medical image analysis method disclosed in this application, the correction parameter includes a parameter associated with a linear function for calculating the first morphological value based on the second morphological value, and obtaining the corrected morphological value may include obtaining the corrected morphological value based on the linear function including the target morphological value and the parameter.

[0082] According to the medical image analysis method disclosed in this application, the first scan condition and the second scan condition may be associated with at least one of the setting parameters associated with the resolution of the medical image of the medical image acquisition device, the magnetic field strength associated with the resolution of the medical image acquisition device, the manufacturer of the medical image acquisition device, and the form of the magnetic field generated by the medical image acquisition device.

[0083] According to the medical image analysis method disclosed in this application, the target medical image is obtained from a first subject having a first characteristic, the correction parameters are obtained from the first medical image and the second medical image obtained from a second subject having the first characteristic, and the first characteristic may be related to the age and sex of the subject.

[0084] According to the medical image analysis method disclosed in this application, the segmentation may be performed using a neural network provided to acquire multiple regions corresponding to multiple elements based on the target medical image.

[0085] The medical image analysis method disclosed in this application further includes converting the target medical image acquired under the second scan conditions into a medical image corresponding to the image acquired under the first scan conditions, wherein the segmentation may be performed based on the converted medical image.

[0086] According to the medical image analysis method disclosed in this application, when the first scan conditions and the second scan conditions are related to magnetic field strength, the correction parameters include a set of parameters for converting the second morphological values ​​related to a target element, obtained from the second medical image acquired under the second magnetic field strength, to the first morphological values ​​related to a target element, obtained from the first medical image acquired under the first magnetic field strength, and when the target medical image is acquired under the second magnetic field strength, the corrected morphological values ​​may be obtained based on the target morphological values ​​and the set of parameters.

[0087] The medical image analysis device disclosed in this application comprises an image acquisition unit that acquires a target medical image, and a controller that provides medical image analysis information based on the target medical image, wherein the controller may be configured to acquire a correction parameter calculated based on the correlation between a first morphological value acquired from a first medical image acquired under first scan conditions and associated with a target element, and a second morphological value acquired from a second medical image acquired under second scan conditions and associated with the target element, acquire a target medical image acquired under the second scan conditions, perform segmentation of the target medical image into a plurality of regions corresponding to a plurality of elements including the target element to acquire a target region associated with the target element from the target medical image, acquire a target morphological value associated with the target element based on voxel data corresponding to the target region, acquire a morphological value corrected based on the target morphological value and the correction parameter, and output a morphological index based on the corrected morphological value.

[0088] According to the medical image analysis apparatus disclosed in this application, the correction parameter includes a parameter for calculating the first morphological value based on the second morphological value, and the controller can obtain the corrected morphological value, which is a morphological estimate of the target element under the first scan conditions, based on the target morphological value and the correction parameter.

[0089] According to the medical image analysis apparatus disclosed in this application, the correction parameter includes a parameter associated with a linear function for calculating the first morphological value based on the second morphological value, and the controller may be configured to obtain the corrected morphological value based on the target morphological value and the linear function including the parameter.

[0090] According to the medical image analysis apparatus disclosed in this application, the first scan condition and the second scan condition may be associated with at least one of the setting parameters associated with the resolution of the medical image of the medical image acquisition apparatus, the magnetic field strength associated with the medical image acquisition apparatus, the manufacturer of the medical image acquisition apparatus, and the form of the magnetic field generated by the medical image acquisition apparatus.

[0091] According to the medical image analysis device disclosed in this application, the target medical image is obtained from a first subject having a first characteristic, the correction parameters are obtained from the first medical image and the second medical image obtained from a second subject having the first characteristic, and the first characteristic may be related to the age and sex of the subject.

[0092] According to the medical image analysis device disclosed in this application, the segmentation may be performed using a neural network provided to acquire multiple regions corresponding to multiple elements based on the target medical image.

[0093] According to the medical image analysis apparatus disclosed in this application, the controller may be configured to convert the target medical image acquired under the second scan conditions into a medical image corresponding to the image acquired under the first scan conditions, and the segmentation may be performed based on the converted medical image.

[0094] The medical image analysis device disclosed in this application further comprises an output module that outputs morphological information acquired based on the morphological index and the morphological index database, wherein the morphological information includes percentile information of the morphological index associated with the target object of the target medical image relative to the morphological index database, based on the morphological index database and the morphological index, and the output module may be configured to output the morphological information reflecting the percentile information.

[0095] According to the medical image analysis apparatus disclosed in this application, when the first scan conditions and the second scan conditions are related to magnetic field strength, the correction parameters include a set of parameters for converting the second morphological values ​​obtained from the second medical image acquired under the second magnetic field strength and related to a target element to the first morphological values ​​obtained from the first medical image acquired under the first magnetic field strength and related to a target element, and when the target medical image is acquired under the second magnetic field strength, the corrected morphological values ​​may be obtained based on the target morphological values ​​and the set of parameters.

[0096] The medical image analysis method disclosed in this application includes: acquiring a target medical image; acquiring target scan conditions associated with the target medical image; acquiring target morphological values ​​associated with the target element based on voxel data of a target region corresponding to a target element included in the target medical image; determining a target correction parameter from one or more correction parameters based on the target scan conditions; and outputting a morphological index based on the determined target correction parameter and the target morphological values, wherein determining the target correction parameter may include, if the target scan conditions correspond to the first scan conditions, the target correction parameter being determined by a first correction parameter for correcting a first morphological value associated with the target element acquired under the first scan conditions, and if the target scan conditions correspond to the second scan conditions, the target correction parameter being determined by a second correction parameter for correcting a second morphological value associated with the target element acquired under second scan conditions different from the first scan conditions.

[0097] According to the medical image analysis method disclosed in this application, outputting the morphological index may include: calculating a first corrected morphological index based on the target morphological value and the first correction parameter when the target scan condition corresponds to the first scan condition; calculating a second corrected morphological index different from the first corrected morphological index based on the target morphological value and the second correction parameter when the target scan condition corresponds to the second scan condition; outputting the first corrected morphological index when the target scan condition corresponds to the first scan condition; and outputting the second corrected morphological index when the target scan condition corresponds to the second scan condition.

[0098] The medical image analysis method disclosed in this application may further include obtaining reference scan conditions that serve as a basis for determining the target correction parameters.

[0099] According to the medical image analysis method disclosed in this application, the first correction parameter includes a parameter for calculating a morphological value obtained under the first scan conditions to a morphological estimate under the third scan conditions, the second correction parameter includes a parameter for calculating a morphological value obtained under the second scan conditions to a morphological estimate under the third scan conditions, calculating the first corrected morphological index includes obtaining the first corrected morphological value, which is a morphological estimate of the target element under the second scan conditions, based on the target morphological value and the first correction parameter, and calculating the second corrected morphological index includes obtaining the second corrected morphological value, which is a morphological estimate of the target element under the third scan conditions, based on the target morphological value and the second correction parameter.

[0100] According to the medical image analysis method disclosed in this application, the first correction parameter includes a first parameter set associated with a first linear function for calculating a morphological value obtained under the first scan conditions to a morphological estimate under the third scan conditions, the second correction parameter includes a second parameter set associated with a second linear function for calculating a morphological value obtained under the second scan conditions to a morphological estimate under the third scan conditions, calculating the first corrected morphological index includes obtaining the first corrected morphological value based on the first linear function including the target morphological value and the first parameter set, and calculating the second corrected morphological index includes obtaining the second corrected morphological value based on the second linear function including the target morphological value and the second parameter set.

[0101] According to the medical image analysis method disclosed in this application, the first scan condition, the second scan condition, and the target scan condition can be associated with at least one of the setting parameters associated with the resolution of the medical image of the medical image acquisition device, the magnetic field strength associated with the medical image acquisition device, the manufacturer of the medical image acquisition device, and the form of the magnetic field generated by the medical image acquisition device.

[0102] According to the medical image analysis method disclosed in this application, the target medical image is obtained from an object having a first characteristic, the first correction parameter or the second correction parameter is obtained from the first medical image and the second medical image obtained from the object having the first characteristic, and the first characteristic may be related to the age and sex of the object.

[0103] According to the medical image analysis method disclosed in this application, obtaining the target morphological value includes obtaining a target region associated with the target element from the target medical image by performing segmentation of the target medical image into a plurality of regions corresponding to a plurality of elements including at least the target element, and obtaining the target morphological value based on voxel data corresponding to the target region, wherein the segmentation may be performed using a neural network provided to obtain a plurality of regions corresponding to a plurality of elements based on the target medical image.

[0104] The medical image analysis method disclosed in this application further includes converting the target medical image acquired under the target scan conditions into a medical image corresponding to an image taken under scan conditions other than the target scan conditions, and the segmentation may be performed based on the converted medical image.

[0105] According to the medical image analysis method disclosed in this application, when the first scan condition, the second scan condition, and the target scan condition are related to magnetic field strength, the first correction parameter includes a first set of parameters for correcting the first morphological value associated with the target element acquired under the first magnetic field strength, and the second correction parameter includes a second set of parameters for correcting the second morphological value associated with the target element acquired under the second magnetic field strength, and determining the target correction parameter may include determining the target correction parameter to the first set of parameters when the target scan condition corresponds to the first magnetic field strength, and determining the target correction parameter to the second set of parameters when the target scan condition corresponds to the second magnetic field strength.

[0106] The medical image analysis device disclosed in this application comprises an image acquisition unit that acquires a target medical image and a controller that provides medical image analysis information based on the target medical image, wherein the controller is configured to acquire target scan conditions related to the target medical image, acquire target morphological values ​​related to the target element based on voxel data of a target region corresponding to a target element included in the target medical image, determine a target correction parameter from one or more correction parameters based on the target scan conditions, and output a morphological index based on the determined target correction parameter and the target morphological value, wherein the target correction parameter may be determined by determining a first correction parameter for correcting a first morphological value related to the target element acquired under the first scan conditions when the target scan conditions correspond to the first scan conditions, and by determining a second correction parameter for correcting a second morphological value related to the target element acquired under second scan conditions different from the first scan conditions when the target scan conditions correspond to the second scan conditions.

[0107] According to the medical image analysis device disclosed in this application, the controller may be configured to output the morphological index by calculating a first corrected morphological index based on the target morphological value and the first correction parameter when the target scan condition corresponds to the first scan condition, calculating a second corrected morphological index different from the first corrected morphological index based on the target morphological value and the second correction parameter when the target scan condition corresponds to the second scan condition, outputting the first corrected morphological index when the target scan condition corresponds to the first scan condition, and outputting the second corrected morphological index when the target scan condition corresponds to the second scan condition.

[0108] According to the medical image analysis apparatus disclosed in this application, the first correction parameter includes a parameter for calculating a morphological value obtained under the first scan conditions to a morphological estimate under the third scan conditions, the second correction parameter includes a parameter for calculating a morphological value obtained under the second scan conditions to a morphological estimate under the third scan conditions, and the controller may be configured to obtain the first corrected morphological value, which is the morphological estimate of the target element under the second scan conditions, based on the target morphological value and the first correction parameter, and to obtain the second corrected morphological value, which is the morphological estimate of the target element under the third scan conditions, based on the target morphological value and the second correction parameter, and to calculate the second corrected morphological index.

[0109] According to the medical image analysis apparatus disclosed in this application, the first correction parameter includes a first parameter set associated with a first linear function for calculating a morphological estimate of a morphological value obtained under the first scan conditions under third scan conditions, the second correction parameter includes a second parameter set associated with a second linear function for calculating a morphological estimate of a morphological value obtained under the second scan conditions under third scan conditions, and the controller may be configured to obtain the first corrected morphological value and calculate the first corrected morphological index based on the first linear function including the target morphological value and the first parameter set, and to obtain the second corrected morphological value and calculate the second corrected morphological index based on the second linear function including the target morphological value and the second parameter set.

[0110] According to the medical image analysis apparatus disclosed in this application, the first scan condition and the second scan condition may be associated with at least one of the setting parameters associated with the resolution of the medical image of the medical image acquisition apparatus, the magnetic field strength associated with the medical image acquisition apparatus, the manufacturer of the medical image acquisition apparatus, and the form of the magnetic field generated by the medical image acquisition apparatus.

[0111] According to the medical image analysis apparatus disclosed in this application, the target medical image is obtained from an object having a first characteristic, and the target correction parameter, which includes at least one of the first correction parameter and the second correction parameter, is obtained from the first medical image and the second medical image obtained from the object having the first characteristic, and the first characteristic may be related to the age and sex of the object.

[0112] According to the medical image analysis apparatus disclosed in this application, the controller is configured to obtain target regions associated with the target element from the target medical image by performing segmentation of the target medical image into a plurality of regions corresponding to a plurality of elements including at least the target element, and to obtain the target morphological value based on voxel data corresponding to the target region, wherein the segmentation may be performed using a neural network provided to obtain a plurality of regions corresponding to a plurality of elements based on the target medical image.

[0113] According to the medical image analysis device disclosed in this application, the controller may be configured to convert the target medical image acquired under the target scan conditions into a medical image corresponding to an image taken under scan conditions other than the target scan conditions, and to perform the segmentation based on the converted medical image.

[0114] According to the medical image analysis apparatus disclosed in this application, when the first scan condition, the second scan condition, and the target scan condition are related to magnetic field strength, the first correction parameter includes a first parameter set for correcting the first morphological value related to the target element acquired under the first magnetic field strength, and the second correction parameter includes a second parameter set for correcting the second morphological value related to the target element acquired under the second magnetic field strength, and the controller may be configured to determine the target correction parameter by determining the target correction parameter to the first parameter set when the target scan condition corresponds to the first magnetic field strength, and the target correction parameter to the second parameter set when the target scan condition corresponds to the second magnetic field strength.

[0115] The brain image analysis method disclosed in this application may include: obtaining a first correction parameter for correcting a first morphological value associated with a first brain element; obtaining a second correction parameter for correcting a second morphological value associated with a second brain element different from the first brain element; obtaining a target brain image; performing segmentation of the target brain image into a plurality of brain regions including a first region corresponding to the first brain element, a second region corresponding to the second brain element, and a cranial region, thereby obtaining a third region associated with the internal regions of the first region, the second region, and the cranial region; obtaining a first brain morphological index associated with the first brain element based on voxel data corresponding to the first region, voxel data corresponding to the third region, and the first correction parameter; and obtaining a second brain morphological index associated with the second brain element based on voxel data corresponding to the second region, voxel data corresponding to the third region, and the second correction parameter.

[0116] The brain imaging analysis method disclosed in this application further includes obtaining a reference morphological value associated with the internal region based on voxel data corresponding to the third region, obtaining the first brain morphological index includes obtaining a first target morphological value associated with the first brain element based on voxel data corresponding to the first region, calculating a first corrected morphological value associated with the first brain element based on the first target morphological value and the first correction parameter, and calculating the first brain morphological index based on the first corrected morphological value and the reference morphological value, and obtaining the second brain morphological index may include obtaining a second target morphological value associated with the second brain element based on voxel data corresponding to the second region, calculating a second corrected morphological value associated with the second brain element based on the second target morphological value and the second correction parameter, and calculating the second brain morphological index based on the second corrected morphological value and the reference morphological value.

[0117] According to the brain imaging analysis method disclosed in this application, the first morphological value and the second morphological value are values ​​obtained under first scan conditions, the target brain image is obtained under first scan conditions, the first correction parameter includes a first parameter for calculating a third morphological value associated with the first brain element obtained under second scan conditions different from the first scan conditions based on the first morphological value, the second correction parameter includes a second parameter for calculating a fourth morphological value associated with the second brain element obtained under second scan conditions based on the second morphological value, calculating the first corrected morphological value includes obtaining the first corrected morphological value, which is a morphological estimate of the first brain element under second scan conditions, based on the first target morphological value and the first correction parameter, and calculating the second corrected morphological value may include obtaining the second corrected morphological value, which is a morphological estimate of the second brain element under second scan conditions, based on the second target morphological value and the second correction parameter.

[0118] According to the brain imaging analysis method disclosed in this application, the first correction parameter includes a first parameter set associated with a first linear function for calculating the third morphological value based on the first morphological value, the second correction parameter includes a second parameter set associated with a second linear function for calculating the fourth morphological value based on the second morphological value, calculating the first corrected morphological value includes obtaining the first corrected morphological value based on the first target morphological value and the first parameter set associated with the first linear function, and calculating the second corrected morphological value includes obtaining the second corrected morphological value based on the second target morphological value and the second parameter set associated with the second linear function.

[0119] According to the brain imaging analysis method disclosed in this application, the first region corresponding to the first brain element can be located adjacent to the skull region compared to the second region corresponding to the second brain element.

[0120] According to the brain imaging analysis method disclosed in this application, the first brain element may be associated with an element that performs a first brain function, and the second brain element may be associated with an element that performs a second brain function different from the first brain function.

[0121] According to the brain imaging analysis method disclosed in this application, the first scan condition and the second scan condition can be associated with at least one of the setting parameters associated with the resolution of the brain image of the brain imaging device, the magnetic field strength associated with the brain imaging device, the manufacturer of the brain imaging device, and the form of the magnetic field generated by the brain imaging device.

[0122] According to the brain imaging analysis method disclosed in this application, the target brain image is obtained from a first subject having a first characteristic, the first correction parameter and the second correction parameter are obtained from a first brain image and a second brain image obtained from a second subject having the first characteristic, and the first characteristic may be related to the age and sex of the subject.

[0123] According to the brain imaging analysis method disclosed in this application, the segmentation may be performed using a neural network provided to acquire a plurality of brain regions corresponding to a plurality of brain elements, including the first brain element, the second brain element, and the skull, based on the target brain imaging.

[0124] The brain imaging analysis method disclosed in this application further includes preprocessing the target brain imaging, which includes performing preprocessing on the first region using a first method and performing preprocessing on the second region using a second method different from the first method, wherein the segmentation may be performed based on the preprocessed target brain imaging.

[0125] According to the brain imaging analysis method disclosed in this application, the first corrected morphological value is a volume value associated with the first brain element obtained in the target brain imaging, the second target morphological value is a volume value associated with the second brain element obtained in the target brain imaging, the first brain morphological index is calculated using the first corrected morphological value relative to the reference morphological value, and the second brain morphological index may be calculated using the second corrected morphological value relative to the reference morphological value.

[0126] The brain image analysis device disclosed in this application comprises an image acquisition unit that acquires a target brain image, and a controller that provides brain image analysis information based on the target brain image, wherein the controller may be configured to acquire a first correction parameter for correcting a first morphological value associated with a first brain element, acquire a second correction parameter for correcting a second morphological value associated with a second brain element different from the first brain element, acquire a target brain image, and perform segmentation of the target brain image into a plurality of brain regions including a first region corresponding to the first brain element, a second region corresponding to the second brain element, and a skull region, thereby acquiring a third region associated with the internal region of the first region, the second region, and the skull region, acquire voxel data corresponding to the first region, voxel data corresponding to the third region, and a first brain morphological index associated with the first brain element based on the first correction parameter, and acquire voxel data corresponding to the second region, voxel data corresponding to the third region, and a second brain morphological index associated with the second brain element based on the second correction parameter.

[0127] According to the brain imaging analysis device disclosed in this application, the controller may be configured to acquire a reference morphological value associated with the internal region based on voxel data corresponding to the third region, acquire a first target morphological value associated with the first brain element based on voxel data corresponding to the first region, calculate a first corrected morphological value associated with the first brain element based on the first target morphological value and the first correction parameter, calculate the first brain morphological index based on the first corrected morphological value and the reference morphological value to acquire the first brain morphological index, acquire a second target morphological value associated with the second brain element based on voxel data corresponding to the second region, calculate a second corrected morphological value associated with the second brain element based on the second target morphological value and the second correction parameter, and calculate the second brain morphological index based on the second corrected morphological value and the reference morphological value to acquire the second brain morphological index.

[0128] According to the brain imaging analysis device disclosed in this application, the first morphological value and the second morphological value are values ​​obtained under first scan conditions, the target brain image is obtained under first scan conditions, the first correction parameter includes a first parameter for calculating a third morphological value associated with the first brain element obtained under second scan conditions different from the first scan conditions based on the first morphological value, the second correction parameter includes a second parameter for calculating a fourth morphological value associated with the second brain element obtained under second scan conditions based on the second morphological value, and the controller may be configured to calculate the first corrected morphological value, which is a morphological estimate of the first brain element under second scan conditions, based on the first target morphological value and the first correction parameter, and to calculate the second corrected morphological value, which is a morphological estimate of the second brain element under second scan conditions, based on the second target morphological value and the second correction parameter.

[0129] According to the brain imaging analysis device disclosed in this application, the first correction parameter includes a first parameter set associated with a first linear function for calculating the third morphological value based on the first morphological value, the second correction parameter includes a second parameter set associated with a second linear function for calculating the fourth morphological value based on the second morphological value, calculating the first corrected morphological value includes obtaining the first corrected morphological value based on the first target morphological value and the first parameter set associated with the first linear function, and the controller may be configured to calculate the first corrected morphological value based on the first target morphological value and the first parameter set associated with the first linear function, and to calculate the second corrected morphological value based on the second target morphological value and the second parameter set associated with the second linear function.

[0130] According to the brain imaging analysis device disclosed in this application, the first region corresponding to the first brain element can be located adjacent to the skull region compared to the second region corresponding to the second brain element.

[0131] According to the brain imaging analysis device disclosed in this application, the first brain element may be associated with an element that performs a first brain function, and the second brain element may be associated with an element that performs a second brain function different from the first brain function.

[0132] According to the brain imaging analysis device disclosed in this application, the first scan condition and the second scan condition can be associated with at least one of the setting parameters associated with the resolution of the brain image of the brain imaging device, the magnetic field strength associated with the brain imaging device, the manufacturer of the brain imaging device, and the form of the magnetic field generated by the brain imaging device.

[0133] According to the brain imaging analysis device disclosed in this application, the controller obtains the target brain image from a first subject having a first characteristic, and the first and second correction parameters are obtained from the first and second brain images obtained from a second subject having the first characteristic, and the first characteristic may be related to the age and sex of the subject.

[0134] According to the brain imaging analysis device disclosed in this application, the segmentation may be performed using a neural network provided to acquire a plurality of brain regions corresponding to a plurality of brain elements, including the first brain element, the second brain element, and the skull, based on the target brain image.

[0135] According to the brain imaging analysis device disclosed in this application, the controller preprocesses the target brain image, and the segmentation is performed based on the preprocessed target brain image, and may be configured to perform preprocessing on the first region in a first way and on the second region in a way different from the first way.

[0136] According to the brain imaging analysis device disclosed in this application, the first corrected morphological value is a volume value associated with the first brain element obtained in the target brain image, the second target morphological value is a volume value associated with the second brain element obtained in the target brain image, the first brain morphological index is calculated using the first corrected morphological value relative to the reference morphological value, and the second brain morphological index may be calculated using the second corrected morphological value relative to the reference morphological value.

[0137] The aforementioned objectives, features, and advantages of this application will become even clearer through the following detailed description relating to the attached drawings. However, this application is subject to various modifications and may have various embodiments, and the present invention will be described in detail below with reference to specific embodiments illustrated in the drawings.

[0138] Throughout the specification, the same reference numerals generally indicate the same components. Furthermore, components with the same function within the same conceptual scope shown in the drawings of each embodiment are described using the same reference numerals, and redundant explanations are omitted.

[0139] If a detailed explanation of a known function or configuration related to this application is deemed unnecessary to the summary of this application, such detailed explanation will be omitted. Furthermore, the numbers used in the description of this specification (e.g., 1st, 2nd, etc.) are merely identifiers to distinguish one component from another.

[0140] Furthermore, the suffixes "module" and "part" used for the components in the following embodiments are added or used interchangeably for the purpose of facilitating the creation of the specification and do not have any meaning or role that distinguishes them from one another.

[0141] In the following examples, singular expressions also include plural expressions unless the context clearly indicates otherwise.

[0142] In the following embodiments, terms such as "includes" or "having" mean that the features or components described in the specification are present, and do not preclude the possibility that one or more other features or components may be added.

[0143] In drawings, the size of components may be enlarged or reduced to facilitate explanation. For example, the size and thickness of each component shown in the drawings are arbitrarily shown for explanatory purposes, and the present invention is not necessarily limited to what is shown.

[0144] Where an alternative embodiment is feasible, the order of certain processes may differ from the order described. For example, two processes described consecutively may be executed substantially simultaneously, or in the reverse order of the description.

[0145] In the following embodiments, when components are connected, this includes not only cases where components are directly connected, but also cases where components are indirectly connected through the interposition of other components.

[0146] For example, when components etc. are electrically connected in this specification, this includes not only cases where components etc. are directly electrically connected, but also cases where components etc. are interposed between them and they are indirectly electrically connected.

[0147] The following describes the brain image analysis device, brain image analysis system, and brain image analysis method of this application. Here, brain image analysis may be performed by acquiring multiple brain regions by performing segmentation on brain images, calculating morphological indicators related to the brain, and outputting these to the user.

[0148] This application discloses various techniques to improve the accuracy and reliability of such brain imaging analysis, including checking the quality of brain imaging to improve the reliability of brain imaging analysis results, correcting morphological values ​​and morphological indicators of specific brain regions to improve the accuracy of brain imaging analysis results, calculating morphological values ​​and morphological indicators of specific brain regions from the brain imaging being analyzed to provide personalized analysis results, and providing user-friendly reports by providing only selective information to the user.

[0149] The following describes a medical image analysis method, a medical image analysis apparatus, and a medical image analysis system according to one embodiment of this application.

[0150] Figure 1 is a schematic diagram of a medical image analysis system according to one embodiment of the present application. Referring to Figure 1, the medical image analysis system according to one embodiment of the present application may include an image acquisition device 1000 and an image analysis device 2000.

[0151] The image acquisition device 1000 can acquire an image and transmit it to the image analysis device 2000 via the network.

[0152] For example, the image acquisition device 1000 may be a device for acquiring magnetic resonance imaging (MRI). In this case, the magnetic resonance imaging acquired by the image acquisition device 1000 can be transmitted to the image analysis device 2000 via a network.

[0153] As another example, the image acquisition device 1000 may be a device for acquiring computed tomography (CT) images. In this case, the computed tomography images acquired by the image acquisition device 1000 can be transmitted to the image analysis device 2000 via a network.

[0154] Another example is that the image acquisition device 1000 may be a device for acquiring images obtained by radiography. In this case, the radiography images acquired by the image acquisition device 1000 can be transmitted to the image analysis device 2000 via a network.

[0155] Another example is that the image acquisition device 1000 could be a device for acquiring images obtained by positron emission tomography (PET).

[0156] However, the image acquisition device 1000 described above is merely an example and is not limited thereto. It should be interpreted to include any appropriate device or system used in medical imaging.

[0157] The image acquired by the image acquisition device 1000 may be a two-dimensional image. In this case, the image may include pixel information related to the coordinates, color, and intensity of pixels.

[0158] The image acquired by the image acquisition device 1000 may be a three-dimensional image. In this case, the image may include pixel information related to the coordinates, color, and intensity of voxels.

[0159] The image acquired by the image acquisition device 1000 may include information related to the alignment of the image. For example, the image acquisition device 1000 can acquire data (ijk) related to the orientation of the captured image, taking into account the orientation of the reference coordinate axis (RAS) of the object 100. Specifically, the image acquisition device can acquire data (ijk) related to the orientation of the captured image, taking into account information (xyz) for the coordinate axes of the image acquisition device and information (RAS) for the reference coordinate axis of the object 100.

[0160] The image acquisition device 1000 can acquire data related to the magnetic field strength of the image, data related to the manufacturer of the image device, data related to the setting parameters of the image device, data related to the object being imaged, and so on.

[0161] Data related to magnetic field strength may include data related to the strength of the magnetic field applied to a subject when acquiring medical images using a magnetic field (e.g., MRI). For example, if an image of a subject is acquired under a magnetic field strength of 1.5T (Tesla), the data corresponding to "1.5T" can be considered data related to magnetic field strength. However, it is not limited to 1.5T; data corresponding to magnetic field strengths commonly used to acquire medical images, such as 3T, 7T, and 8T, can also be considered data related to magnetic field strength.

[0162] The setting parameters of an imaging device may include parameters that can be adjusted or controlled by the imaging device in order to acquire medical images. For example, the setting parameters of an imaging device may include the repetition time (TR), which is the time between consecutive pulse sequences applied to the same slice; the time to echo (TE), which is the time between the transmission of an RF pulse and the reception of an echo signal; the time constant (T1), which is related to the rate at which excited protons return to equilibrium; the time constant (T2), which is related to the rate at which excited protons reach equilibrium or move out of phase with each other; the proton density (PD); or any combination thereof.

[0163] However, this should be interpreted to include any parameters related to magnetic field characteristics, not limited to the examples given above. Furthermore, it should be interpreted to include any parameters related to imaging devices that utilize CT or radiation (e.g., X-ray) imaging methods, not limited to acquiring target images using magnetic fields.

[0164] At this time, the data described above is processed as metadata within the acquired image and transmitted to the image analysis device 2000, or it can be transmitted to the image analysis device 2000 separately from the image.

[0165] The image acquisition device 1000 may be a device for acquiring medical images such as MRI, CT, and X-ray.

[0166] The image acquisition device 1000 can acquire images under various scanning conditions.

[0167] For example, if the imaging device is an MRI, an image can be acquired under scanning conditions with a magnetic field strength of 1.5 Tesla (hereinafter referred to as T). An image can also be acquired under scanning conditions with a magnetic field strength of 3T. Furthermore, an image can be acquired under scanning conditions with a magnetic field strength of 7T or 8T.

[0168] In another example, if the image acquisition device 1000 is an MRI, then, as described above, images can be acquired under scan conditions of setting parameters consisting of TR (Repetition Time), TE (Time to Echo), proton density (PD), or a time constant (T1) related to the rate at which excited protons return to equilibrium, a time constant (T2) related to the rate at which excited protons reach equilibrium or move out of phase with each other, or any combination thereof.

[0169] In another example, the image acquisition device 1000 may be a device associated with a CT scanner. In this case, an image can be acquired under scanning conditions related to setting parameters such as voltage, current, exposure time, scanning time, projection, and any combination thereof, which can be set on the CT scanner.

[0170] In another example, the image acquisition device 1000 may be an X-ray-related device. In this case, an image can be acquired under scan conditions related to the setting parameters that can be set on the X-ray imaging device, such as tube voltage, tube current, exposure time, distance between the X-ray tube device and the detector (SID), angle of the X-ray tube support device, and any combination thereof.

[0171] Furthermore, the image may be acquired differently depending on the characteristics of the applicator who irradiates the image. In this case, information regarding the applicator's characteristics can be parameterized. The parameters regarding the applicator's characteristics may also be used as one of the considerations in relation to correcting morphological values ​​and morphological indicators considering the scan conditions described later. In other words, morphological values ​​and morphological indicators can be corrected according to the applicator's characteristics. More specifically, the image analysis device 2000 according to one embodiment can acquire applicator identification information in relation to the scan conditions. The image analysis device 2000 may also be implemented to acquire correction parameters for correcting morphological values ​​and morphological indicators based on the applicator's identification information. For example, the image analysis device 2000 can obtain correction parameters (or correction parameters for estimating the morphological values ​​calculated from the target image obtained from the first practitioner to be associated with the morphological values ​​obtained from the second practitioner) based on the correlation between the first morphological values ​​calculated from the first image taken by the first practitioner and the second morphological values ​​calculated from the second image taken by the second practitioner.

[0172] In another example, the image acquisition device 1000 may be an MRI or CT scanner using positron emission tomography (PET). When using PET imaging, drugs related to radiopharmaceuticals may be used. For example, drugs (or tracers) such as 18F-florbetapir, 18F-florbetaben, 18F-flutemetamol, and 18F-florapronol may be used to measure amyloid beta. Drugs (or tracers) such as 18F-flortaucipir may be used to measure tau protein. The drugs used may vary depending on the manufacturer of the image acquisition device. In this case, information about the drugs used may be parameterized. The parameters for the drugs used may be used as scan conditions for the image acquisition device and may be used as one of the considerations in relation to correcting morphological values ​​and morphological indicators considering the scan conditions described later. In other words, morphological values ​​and morphological indicators can be corrected based on the manufacturer of the image acquisition device and the chemicals used. More specifically, the image analysis device 2000 can acquire information on the manufacturer of the image acquisition device and the chemicals used, and can be implemented to acquire correction parameters for correcting morphological values ​​and morphological indicators based on this information.For example, the image analyzer 2000 can obtain a first correction parameter (or a correction parameter for estimating the morphological value calculated from a target image acquired using the second chemical) based on the correlation between a first morphological value calculated from a first image acquired using the first chemical (e.g., 18F-florbetapir) and a second morphological value calculated from a second image acquired using the second chemical (e.g., 18F-florbetaben). Furthermore, the image analyzer 2000 can obtain a second correction parameter (or a correction parameter for estimating the morphological value calculated from a target image acquired using the third drug to the morphological value acquired using the second drug) based on the correlation between the second morphological value calculated from a second image acquired using the second drug and the third morphological value calculated from a third image acquired using the third drug. This correction parameter may differ from the second correction parameter.

[0173] Furthermore, the image acquisition device 1000 can acquire images under scan conditions that include at least one or more setting parameters. In other words, images may be acquired under scan conditions based on various combinations of the setting parameters described above.

[0174] For example, an image may be acquired under a first scan condition where the set parameters are a first combination. Alternatively, an image may be acquired under a second scan condition where the set parameters are a second combination and the magnetic field strength is a first strength. However, the images acquired by the image acquisition device 1000 may be images acquired under various scan conditions with various combinations of set parameters.

[0175] Furthermore, the image acquisition device 1000 may be composed of multiple image acquisition devices 1000.

[0176] In this case, the multiple image acquisition devices 1000 may be devices manufactured by different manufacturers. Images acquired by devices manufactured by different manufacturers may have different characteristics, such as brightness and intensity, even if the scan conditions and setting parameters are the same. Therefore, even when acquiring medical images of the same subject, the morphological indicators based on the medical images may differ depending on the manufacturer's device.

[0177] Therefore, the correction parameter acquisition device 2400 according to one embodiment of this application, described later, is required to perform a correction parameter acquisition operation to control scan conditions and setting parameters, or variables provided by the image device manufacturer.

[0178] Images acquired by the image acquisition device 1000 may include information related to the anatomical structure of a specific part of the body. Furthermore, the specific part of the body can correspond to any part where medical imaging can be utilized. For the sake of explanation, the drawings and specification described later will focus on images related to the brain, but this is merely illustrative, and the various embodiments disclosed in this application can be applied to any suitable part of the body other than the brain (e.g., lungs, breasts, heart, joints, blood vessels, etc.).

[0179] On the other hand, the image acquisition device 1000 according to one embodiment of this application can be implemented in the form of a server. In this case, the server may be configured to store medical images and information related to the medical images. The server may also be configured to modify or process the medical images and information related to the medical images.

[0180] Furthermore, medical images may be stored in the memory or server of the image analysis device 2000 and used to calculate correction parameters, perform quality control, and output analysis results. This will be discussed in more detail below.

[0181] In the following, an image analysis apparatus 2000, a learning device 2200, a correction parameter acquisition device 2400, and an output device 2600 for image analysis according to one embodiment of the present application will be described with reference to Figure 2. Figure 2 is a schematic diagram of the image analysis apparatus 2000, a learning device 2200, a correction parameter acquisition device 2400, and an output device 2600 for medical image analysis according to one embodiment of the present application.

[0182] The image analysis device 2000 can perform operations such as segmenting images acquired from the image acquisition device 1000 using an artificial neural network learned by the learning device 2200, and calculating morphological indicators of target elements contained within the images.

[0183] The learning device 2200 can use multiple image datasets to train neural network models for image segmentation or for image quality control.

[0184] The correction parameter acquisition device 2400 can use multiple image datasets and data on scan conditions to calculate correction parameters for correcting morphological values ​​and morphological indices associated with the target elements of the target image (or target image) acquired by the image analysis device 2000.

[0185] The image analysis device 2000, learning device 2200, and correction parameter acquisition device 2400 shown in Figure 2 may be implemented to send and receive data from each other using any communication method.

[0186] For example, the image analysis device 2000, the learning device 2200, and the correction parameter acquisition device 2400 may be implemented to share a server.

[0187] As shown in Figure 2, the image analysis device 2000, the learning device 2200, and the correction parameter acquisition device 2400 are provided as separate devices. However, this is merely illustrative, and the image analysis device 2000, the learning device 2200, and / or the correction parameter acquisition device 2400 could also be implemented in a single device. Alternatively, some of the image analysis device 2000, the learning device 2200, and the correction parameter acquisition device 2400 could be provided as separate devices, while the remaining devices could be implemented in a single device.

[0188] The configuration of an image analysis apparatus 2000 according to one embodiment of this application will be described below with reference to Figure 3. Figure 3 is a block diagram relating to an image analysis apparatus 2000 according to one embodiment of this application.

[0189] An image analysis apparatus 2000 according to one embodiment of this application may include a first communication module 2010, a first memory 2020, and a first controller 2030.

[0190] The first communication module 2010 can communicate with any external device, including the image acquisition device 1000, the learning device 2200, the correction parameter acquisition device 2400, and the output device 2600 described later. In other words, the image analysis device 2000 can send and receive images from the image acquisition device 1000, or send and receive data with external devices, including the learning device 2200, the parameter acquisition device 2400, the output device 2600 described later, a relay, and a server, via the first communication module 2010.

[0191] For example, the image analysis device 2000 can receive images acquired from the image acquisition device 1000, information about the neural network model learned from the learning device 2200, and information related to the correction parameters calculated from the correction parameter acquisition device 2400 via the first communication module 2010. In another example, the image analysis device 2000 can transmit information related to the image scanning conditions to the correction parameter acquisition device 2400, or transmit information related to the analysis results to the output device 2600 via the first communication module 2010. In yet another example, the image analysis device 2000 can connect to the internet via the first communication module 2010 and upload various data related to the image, information related to the scanning conditions, and information related to the analysis results.

[0192] The first communication module 2010 is broadly divided into wired and wireless types. Since the wired and wireless types each have their own advantages and disadvantages, the image analysis device 2000 may be equipped with both wired and wireless types simultaneously in some cases.

[0193] In the case of wired connections, LAN (Local Area Network) or USB (Universal Serial Bus) communication are typical examples, but other methods are also possible.

[0194] Furthermore, in the case of wireless types, communication methods such as Bluetooth® and Zigbee, which belong to the Wireless Personal Area Network (WPAN) series, can be primarily used. However, wireless communication protocols are not limited by this, and wireless communication modules can also use communication methods such as Wi-Fi, which belong to the Wireless Local Area Network (WLAN) series, or other known communication methods.

[0195] The first memory 2020 can store various types of information. Various types of data can be stored in the first memory 2020 temporarily or semi-permanently. Examples of the first memory 2020 include hard disk drives (HDD), solid state drives (SSD), flash memory, read-only memory (ROM), and random access memory (RAM).

[0196] The first memory 2020 can be provided either as an internal component of the image analysis device 2000 or as a detachable component. The first memory 2020 can store various data necessary for the operation of the image analysis device 2000, including the operating system (OS) for driving the image analysis device 2000 and programs for operating each component of the image analysis device 2000. For example, the first memory 2020 can store various data related to images, information related to scan conditions, and information related to analysis results.

[0197] The first controller 2030 can control the overall operation of the image analyzer 2000. For example, the first controller 2030 can load and execute a program for the operation of the image analyzer 2000 from the first memory 2020.

[0198] The first controller 2030 can be implemented as a CPU (Central Processing Unit) or similar device through hardware, software, or a combination thereof. In terms of hardware, it can be provided in the form of an electronic circuit that processes electrical signals and performs control functions, and in terms of software, it can be provided in the form of a program or code that drives the hardware circuit.

[0199] On the other hand, referring again to Figure 3, the image analysis apparatus 2000 according to one embodiment of the present application may include an input module 2040 and an output module 2050.

[0200] At this time, the image analysis device 2000 can use the input module 2040 and the output module 2050 to acquire user input and output information corresponding to the user input. For example, the image analysis device 2000 can use the input module 2040 to acquire user input requesting data acquisition, user input instructing preprocessing, user input related to image segmentation, and user input for reference scan conditions related to the calculation of morphological indicators, and output the corresponding information via the output module 2050.

[0201] For example, the user can input conditions and settings related to the analysis of the image analyzer 2000 via the input module 2040.

[0202] For example, the user can set correction parameters and associated reference scan conditions for correcting morphological values ​​and morphological indicators acquired from the target image via the input module 2040. In this case, the image analyzer 2000 can perform correction of morphological values ​​and morphological indicators based on the reference scan conditions received from the input module 2040.

[0203] The input module 2040 can be implemented in various forms, such as a mouse, keyboard, or touchpad.

[0204] The output module 2050 can be provided to output notifications and image analysis results from the image analysis operation of the image analysis device 2000.

[0205] For example, when the image analysis device 2000 performs an operation to check the quality of an image, a notification window indicating the presence of artifacts in the target image can be provided via the output module 2050.

[0206] In another example, when the image analysis device 2000 performs an operation to check the quality of an image, a notification window may be provided via the output module 2050 to select whether to perform an image analysis if serious artifacts are present in the target image.

[0207] In another example, if the image analysis device 2000 performs a segmentation operation on a target image, the segmentation results can be provided via the output module 2050.

[0208] In another example, once the image analysis device 2000 has finished analyzing the target image, the analysis results of the target image can be provided via the output module 2050.

[0209] The output module 2050 can be implemented in various forms, such as a display.

[0210] Furthermore, the image analysis device 2000 may further include a user interface for acquiring user input via the input module 2040 and outputting information corresponding to the user input via the output module 2050.

[0211] Figure 3 shows that an image analysis device 2000 according to one embodiment of this application includes an input module 2040 and an output module 2050. However, this is merely illustrative, and an image analysis device 2000 in which the input module 2040 and output module 2050 are omitted can also be provided.

[0212] On the other hand, the image analysis device 2000 according to one embodiment of this application can be implemented in the form of a server. In this case, the server may be configured to store medical images and related information transmitted from the image acquisition device 1000. The server can also be implemented to modify or process the medical images and related information transmitted from the image acquisition device 1000.

[0213] Furthermore, the server for the image analysis device 2000 may be implemented separately from the server for the image acquisition device 1000, but is not limited to this; the servers for the image acquisition device 1000 and the image analysis device 2000 may be implemented in a single form. In other words, the image acquisition device 1000 and the image analysis device 2000 can be implemented in a form that shares a common server.

[0214] On the other hand, referring again to Figure 2, the image analysis apparatus 2000 according to one embodiment of the present application may be implemented to communicate with the output device 2600.

[0215] An output device 2600 according to one embodiment of this application may be configured to receive the analysis results from the image analysis device 2000 and output the image analysis results to the user in a visual format.

[0216] As one embodiment, the output device 2600 according to one embodiment of this application may be implemented to receive image quality and related information from the image analysis device 2000 and output it to the user. For example, the output device 2600 may be implemented to output indicators related to image quality and comments related to the reliability of the image analysis results related to image quality.

[0217] As one embodiment, the output device 2600 according to one embodiment of this application may be configured to receive analysis results related to morphological indicators from the image analysis device 2000 and output them to the user.

[0218] For example, the output device 2600 may be designed to output to the user the results of a comparative analysis of the analysis results related to the morphological indicators of the target body with statistical data of a comparison target population. In this case, the results of the comparative analysis with the statistical data of the comparison target population can be processed and output as statistical information in any appropriate form, such as a graph. The output device 2600 may also be designed to provide analysis results related to the morphological indicators of multiple parts of the target body together.

[0219] In another example, the output device 2600 can be provided to output analysis results related to morphological indicators of the target body, while also outputting information in a visual form related to where the body parts associated with the morphological indicators correspond.

[0220] As one embodiment, the output device 2600 according to one embodiment of this application may be implemented to receive the image segmentation result from the image analysis device 2000 and output it to the user. For example, the segmentation result of the image analysis device 2000 may be the labeled result within the image. In this case, the image analysis device 2000 or the output device 2600 can visually process the image based on the anatomical structure of the body based on the labeled result, and the output device 2600 may be implemented to output an image in which the anatomical structure of the body is visually divided to the user.

[0221] As one embodiment, the output device 2600 according to one embodiment of this application may be implemented to receive a report from the image analysis device 2000 and output it to the user.

[0222] The configuration and operation of the output device 2600 according to one embodiment of this application will be described below with reference to Figure 4. Figure 4 is a block diagram of the output device according to one embodiment of this application.

[0223] An output device 2600 according to one embodiment of this application may include a second communication module 2610, a second memory 2620, and a second controller 2630.

[0224] The second communication module 2610 can communicate with any external device, including the image acquisition device 1000 and the image analysis device 2000. In other words, the output device 2600 can send and receive images from the image acquisition device 1000 and send and receive image analysis results from the image analysis device 2000 via the second communication module 2610. Furthermore, the output device 2600 can send and receive data with any external device, including a repeater, via the second communication module 2610.

[0225] For example, the output device 2600 can receive, via the second communication module 2610, images acquired from the image acquisition device 1000, data related to the quality of images acquired from the image analysis device 2000 and the reliability of the analysis results, data related to image segmentation results, and data such as morphological indicators. In another example, the output device 2600 can transmit data related to user input received from the input module 2640 (described later) to the image analysis device 2000 via the second communication module 2610, or transmit data processed by the output device 2600 to the image analysis device 2000. In yet another example, the output device 2600 can connect to the internet via the second communication module 2610 and upload data related to user input and data processed on the output device 2600.

[0226] The second communication module 2610 is broadly divided into wired and wireless types. Since the wired and wireless types each have their own advantages and disadvantages, the image analysis device 2000 may be equipped with both wired and wireless types simultaneously in some cases.

[0227] In the case of wired connections, LAN (Local Area Network) or USB (Universal Serial Bus) communication are typical examples, but other methods are also possible.

[0228] Furthermore, in the case of wireless types, communication methods such as Bluetooth® and Zigbee, which belong to the Wireless Personal Area Network (WPAN) series, can be primarily used. However, wireless communication protocols are not limited by this, and wireless communication modules can also use communication methods such as Wi-Fi, which belong to the Wireless Local Area Network (WLAN) series, or other known communication methods.

[0229] The second memory 2620 can store various types of information. Various types of data can be stored in the second memory 2620 temporarily or semi-permanently. Examples of the second memory 2620 include hard disk drives (HDD), solid state drives (SSD), flash memory, read-only memory (ROM), and random access memory (RAM).

[0230] The second memory 2620 may be provided in a form that is built into the output device 2600 or in a detachable form. The second memory 2620 may store various data necessary for the operation of the output device 2600, including an operating system (OS) for driving the output device 2600 and programs for operating each component of the output device 2600. For example, the second memory 2620 may store various data related to the image, information related to the analysis results, and user input data.

[0231] The second controller 2630 can control the overall operation of the output device 2600. For example, the second controller 2630 can load and execute a program for the operation of the output device 2600 from the second memory 2620.

[0232] The second controller 2630 can be implemented as a CPU (Central Processing Unit) or similar device through hardware, software, or a combination thereof. In terms of hardware, it can be provided in the form of an electronic circuit that processes electrical signals and performs control functions, and in terms of software, it can be provided in the form of a program or code that drives the hardware circuit.

[0233] On the other hand, referring again to Figure 4, the output device 2600 according to one embodiment of the present application may include an input module 2640 and an output module 2650.

[0234] At this time, the output device 2600 can acquire user input using the input module 2640 and the output module 2650, and output information corresponding to the user input.

[0235] For example, the output device 2600 can use the input module 2640 to acquire user input requesting data acquisition, user input instructing preprocessing, user input related to image segmentation, and user input related to reference scan conditions related to the calculation of morphological indicators, and output the corresponding information via the output module 2650.

[0236] In this case, the output device 2600 may further include a user interface for acquiring user input via the input module 2640 and outputting information corresponding to the user input via the output module 2650.

[0237] In one embodiment, the user can input conditions and settings related to the output of the output device 2600 via the input module 2640, or select a portion of the multiple output data.

[0238] As an example, the user can set correction parameters and associated reference scan conditions for correcting morphological values ​​and morphological indices acquired from the target image via the input module 2640. In this case, the output device 2600 may be configured to output analysis results corresponding to the corrected morphological values ​​and corrected morphological indices based on the reference scan conditions (e.g., magnetic field strength, manufacturer, setting parameters of the imaging device, etc.) received from the input module 2640.

[0239] For example, if a user wants to change the magnetic field strength and be provided with image analysis results corresponding to the changed magnetic field strength, the scan conditions for the magnetic field strength can be input via the input module 2640. In this case, the output device 2600 can be implemented to output image analysis results corrected based on the scan conditions for the magnetic field strength input by the user. In this case, the output device 2600 can transmit input data related to the scan conditions for the magnetic field strength input by the user to the image analysis device 2000, and the image analysis device 2000 can correct the image analysis results based on the correction parameters obtained based on the received user input and transmit them to the output device 2600. Of course, the output device 2600 can be implemented to receive correction parameters from the image analysis device 2000, correct the image analysis results, and output them.

[0240] In relation to this, more specific details will be described later in connection with Figures 66 and 67.

[0241] However, the above is merely illustrative and not limited to it. The output device 2600 can be implemented to receive user input via the input module 2640 in order to output any form of information that the user wishes to provide.

[0242] The input module 2640 may be implemented in various forms, such as a mouse, keyboard, or touchpad.

[0243] The output module 2650 may be provided to output, such as the image analysis result received from the image analysis apparatus 2000.

[0244] For example, the output apparatus 2600 can output, via the output module 2650, information on whether there is an artifact in the target image received from the image analysis apparatus 2000 and the reliability of the analysis result thereof in text or any appropriate visual form.

[0245] According to another example, the output apparatus 2600 can output, via the output module 2650, the result of segmentation of the target image received from the image analysis apparatus 2000. At this time, the segmentation result may be realized to output by separating the anatomical structure of the body in a visual graphic form.

[0246] Also, in another example, the output apparatus 2600 can output, via the output module 2650, the analysis result of the target image received from the image analysis apparatus 2000. At this time, the analysis result of the target image may be output in the form of a report. Further, the morphological values and morphological indexes of the anatomical structure included in the analysis result of the target image may be realized to be output using statistical techniques such as graphs. Also, the morphological values and morphological indexes of the anatomical structure included in the analysis result of the target image may be realized to be output using statistical techniques such as graphs in comparison with the statistical data of the comparison target population.

[0247] The output module 2650 may be realized in various forms such as a display.

[0248] According to the above description, the image analysis apparatus 2000 and the output apparatus 2600 have been described as separate apparatuses. However, this is merely an example, and the image analysis apparatus 2000 and the output apparatus 2600 may be realized as a single apparatus.

[0249] The image analysis device 2000 can receive the medical images transmitted from the image acquisition device 1000 and perform pre-processing such as alignment of the medical images, normalization of the brightness or intensity of the medical images, and noise removal.

[0250] In addition, the image analysis device 2000 can receive the medical images transmitted from the image acquisition device 1000 and perform segmentation of the medical images. At this time, the segmentation of the medical images according to an embodiment of the present application may be performed using a learned neural network model.

[0251] The image analysis device 2000 according to an embodiment of the present application can perform an operation of controlling the quality of the medical images. Therefore, according to the image analysis device 2000 of an embodiment of the present application, the reliability of the medical image analysis result can be improved.

[0252] The image analysis device 2000 according to an embodiment of the present application can perform an operation of calculating morphological indexes based on the medical images acquired from the subject. Therefore, according to the image analysis device 2000 of an embodiment of the present application, accurate morphological indexes can be calculated in relation to the anatomical structure of the subject's body. There is an advantageous effect that this can provide an accurate and objective index for the disease of the subject. In particular, since the brain image related to the subject is not aligned with the standard brain model and the image segmentation is performed to calculate the morphological index, a diagnostic assistance index related to the personalized brain disease can be provided.

[0253] An image analysis device 2000 according to one embodiment of this application can perform operations to correct morphological values ​​and morphological indicators by taking into account the scan conditions or body parts in which medical images are acquired. Therefore, according to the image analysis device 2000 according to one embodiment of this application, accurate morphological indicators can be calculated in relation to the anatomical structure of the target body. In other words, although there may be some errors in the morphological values ​​calculated depending on the scan conditions and body parts, the image analysis device 2000 according to one embodiment of this application can correct the morphological values ​​and morphological indicators that were calculated firsthand by taking into account the scan conditions and body parts. This has the advantageous effect of providing accurate and objective indicators for diseases of the target body.

[0254] An image analysis device 2000 according to one embodiment of this application can perform the operation of assigning priority to various image analysis results according to the state of the subject or the diagnostic field, and outputting image analysis results selected based on the priority. Therefore, according to the image analysis device 2000 according to one embodiment of this application, it is possible to selectively provide the user with the necessary indicator information from among the various indicator information obtained through the analysis of the target medical image, thereby increasing ease of use.

[0255] The following describes in more detail some operations performed by one embodiment of the image analysis device 2000.

[0256] For the sake of clarity, the following explanation will focus on examples of analyzing images related to the brain. However, the various examples disclosed in this application are not limited to the brain and can be applied to medical images of any appropriate body part other than the brain.

[0257] Furthermore, in the following, the terms medical image, brain image, object image (or target image), etc., will be used interchangeably, but this is solely for the sake of explanation. Medical image, brain image, object image (or target image) should be interpreted as referring to images analyzed by both image analysis devices (2000, 3000, 4000).

[0258] Furthermore, the following may be performed using an image analysis apparatus (2000, 3000, 4000) according to one embodiment of this application. The reference numerals indicating the image analysis apparatus are simply used to distinguish and explain the operation of the image analysis apparatus for the sake of explanation, and the reference numerals do not limit the image analysis apparatus.

[0259] An image analysis device 2000 according to one embodiment of this application can acquire brain images and information related to brain images.

[0260] Specifically, the image analysis device 2000 can acquire brain images from the image acquisition device 1000. Furthermore, the image analysis device 2000 can acquire information related to brain images from the image acquisition device 1000 or any external device (e.g., a server).

[0261] More specifically, the image analysis device 2000 can acquire brain images and data related to those brain images from the image acquisition device 1000 via the first communication module 2010.

[0262] In this case, the brain image format can be various medical image formats. For example, the brain image could be DICOM, Nifti (Neuroimaging Ingormatics Technology Initiative), or any other appropriate format.

[0263] The data associated with brain images may include data contained within the brain images themselves, data related to the scan conditions under which the brain images were acquired, and data relating to the subject of the brain images.

[0264] In this case, the data included in the brain image may be data related to pixels or voxels contained within the brain image, data related to the orientation of the brain image, and arbitrary structured metadata for the brain image.

[0265] In particular, data related to the scan conditions under which the brain image is acquired, or data concerning the subject of the brain image, may be structured as metadata for the brain image.

[0266] On the other hand, data related to the scan conditions under which brain images are acquired may include data related to the magnetic field strength of the image acquisition device 1000, the setting parameters of the image acquisition device 1000, or the manufacturer of the image acquisition device.

[0267] Since morphological values ​​obtained from brain images are affected by scanning conditions, the image analysis device 2000 according to one embodiment of this application can acquire data related to the scanning conditions and perform corrections to the morphological values. This will be described in detail later in relation to Figures 58 to 67.

[0268] The data for the subject of the brain image may include personal information and medical information about the subject of the target image analyzed by the image analysis device 2000. For example, the data for the subject of the brain image may include personal information related to the subject's gender, age, etc., as well as various medical information related to questionnaire surveys related to brain diseases (e.g., dementia, Alzheimer's disease, depression, stroke, etc.) or information related to underlying diseases.

[0269] In particular, since gender and age are important variables in brain diseases, data related to the gender and age of the subject can serve as a basis for the image analysis apparatus 2000 to determine or acquire correction parameters. For example, when the target image to be analyzed is obtained from a first subject having a first characteristic related to gender and age, the image analysis apparatus 2000 can determine correction parameters for correcting the morphological values and morphological indicators obtained from the target image based on the correction parameters obtained from the first brain image and the second brain image obtained from a second subject having the first characteristic. For example, the image analysis apparatus 2000 can determine correction parameters for correcting the morphological values and morphological indicators obtained from the target image based on the correction parameters obtained from a second subject having an age similar to that of the first subject of the target image and the same gender.

[0270] Alternatively, data related to the gender and age of the subject can be considered when outputting the relative percentile of the morphological indicators of the subject according to gender or age by utilizing statistical techniques.

[0271] In addition, the image analysis apparatus 2000 according to an embodiment of the present application can acquire information related to operations related to brain image analysis.

[0272] Specifically, the image analysis apparatus 2000 can acquire information related to a template related to the brain for preprocessing or aligning the brain image from an arbitrary external device. This will be described in detail later in relation to FIGS. 5, 6, and 48.

[0273] In addition, the image analysis apparatus 2000 can acquire information related to a brain atlas serving as a reference for segmenting the brain image from an arbitrary external device.

[0274] For example, information related to brain atlases may be information related to atlases related to brain structure. For instance, atlases related to brain structure may include Automated Anatomical Labeling (Tzourio-Mazoyer 2002), Desikan-Killiany Atlas (Desikan 2006), Destrieux Atlas (Destrieux 2010), Harvard-Oxford cortical / subcortical atlases (Makris 2006), MICCAI 2012 Multi-Atlas Labeling Workshop and Challenge (Neuromorphometrics), Hammersmith atlas (Hammers 2003, Gousias 2008, Faillenot 2017), HCP MMP 1.0 (Glasser 2016), JuBrain / Juelich histological atlas (Eickhoff 2005), or MarsAtlas (Auzias 2016).

[0275] As another example, information associated with a brain atlas may be information associated with an atlas related to brain function. For example, atlases related to brain function include: Mindboggle 101 (Klein 2012), Cortical Area Parcellation from Resting-State Correlations (Gordon 2016), Consensual Atlas of Resting-state Network (CAREN, Doucet 2019), Brainnetome Atlas parcellation (Fan 2016), Local-Global Parcellation of the Human Cerebral Cortex (Schaefer 2018), Human Motor Area Template (Mayka 2005), Sensorimotor Area Tract Template (Archer 2017), AICHA: An atlas of intrinsic connectivity of homotopic areas (Joliot 2015), Yeo 2011 functional parcellations (Yeo 2011), PrAGMATiC (Huth 2016), fMRI-based random parcellations (Craddock 2011), and Voxelwise parcellations. This could be (Lead-DBS), SUIT Cerebellar parcellation (Diedrichsen 2006), or Buckner functional cerebellar parcellation (Buckner 2011).

[0276] The image analysis device 2000 can transmit information about the brain-related brain atlas mentioned above to the learning device 2200, and the learning device 2200 may be implemented to learn a neural network model for brain image segmentation based on the information about the brain atlas.

[0277] However, the information on atlases related to brain structure or brain function mentioned above is merely illustrative, and the image analysis device 2000 may be implemented to acquire any appropriate brain-related atlas information to serve as a basis for training an artificial neural network model for image segmentation in the learning device 2200.

[0278] The image analysis device 2000 can receive user input from the input module 2040 of the image analysis device 2000 or from the input module 2640 of the output device 2600.

[0279] The image analysis device 2000 can acquire user input related to the disease being diagnosed. For example, the image analysis device 2000 can acquire user input related to brain diseases (e.g., dementia, depression, stroke, etc.) that correspond to the disease being diagnosed in relation to the target image.

[0280] The image analysis device 2000 can acquire user input related to the patient. For example, the image analysis device 2000 can acquire user input related to the patient's gender, age, name, etc., in relation to the target image.

[0281] The image analysis device 2000 can acquire user input related to image analysis.

[0282] For example, the image analysis device 2000 can acquire user input related to image preprocessing. For instance, the image analysis device 2000 can acquire user input related to image processing, such as correcting the intensity of the target image or removing noise.

[0283] For example, the image analysis device 2000 can acquire user input related to segmentation. For instance, the image analysis device 2000 can acquire user input to modify label data output from brain atlases and neural network models that are considered for segmentation.

[0284] For example, the image analysis device 2000 can receive user input regarding the scan conditions under which the target image was captured via an input module. Information regarding the scan conditions under which the target image was captured is obtained as metadata of the target image, but it can also be obtained through user input.

[0285] As an example, the image analysis device 2000 can receive user input via an input module for reference scan conditions related to the criteria for correction parameters used to correct morphological values ​​and morphological indicators associated with the brain, which are calculated based on the segmentation results. At this time, the image analysis device 2000 can acquire correction parameters based on the user input for reference scan conditions, etc., and correct morphological values ​​and morphological indicators associated with the brain. This will be described in more detail later in relation to Figures 66 and 67.

[0286] In another example, when the image analysis device 2000 outputs image analysis results, it can receive user input via an input module to select some of the output results or to assign priorities to them. In this case, the image analysis device 2000 may selectively output information on the analysis results based on the user input, or output analysis results that reflect the user's priorities. This will be described in detail later in Figures 68 to 81.

[0287] The data acquired by the image analysis device 2000 may be stored in the first memory 2020 of the image analysis device 2000, or in any external device (e.g., a server) connected to the image analysis device 2000. Alternatively, the data acquired by the image analysis device 2000 may be transmitted to the learning device 2200 or the correction parameter acquisition device 2400. Or, the data acquired by the image analysis device 2000 may be transmitted to the output device 2600 or to any external device (e.g., a server).

[0288] An image analysis apparatus 2000 according to one embodiment of this application can perform image preprocessing. The image analysis apparatus 2000 can perform preprocessing to improve the accuracy of image analysis. The image analysis apparatus 2000 may be provided to perform image preprocessing to derive more accurate segmentation results before performing segmentation operations on the image.

[0289] As an example, the image analysis device 2000 may be provided to perform a format conversion of images acquired from the image acquisition device 1000. Specifically, by unifying the format of the images to be analyzed, the neural network model can be trained more stably and accurately. More specifically, it is more stable and accurate to perform image analysis using images that have the same format as the images used to train the neural network model. Therefore, the image analysis device 2000 according to one embodiment of this application may be provided to perform a format conversion of images acquired from the image acquisition device 1000.

[0290] As an example, an image analysis device 2000 according to one embodiment of this application can perform the operation of converting a first-format image acquired from an image acquisition device 1000 into a second-format image. For example, the format of the image acquired from the image acquisition device 1000 may be the DICOM format, which is commonly used in medical imaging. In this case, it is also possible to calculate image segmentation and morphological indices based on the DICOM format image.

[0291] However, when analyzing brain images, brain imaging systems can relatively easily analyze images using Nifti format brain images or train artificial neural networks. Therefore, an image analysis device 2000 according to one embodiment of this application may be provided to perform an operation to convert the format of images acquired from the image acquisition device 1000 to Nifti format.

[0292] However, the above-mentioned formats are merely examples, and the image analysis device 2000 may be provided to perform conversion operations to any appropriate format other than the Nifti format as needed. Furthermore, the format of the image acquired from the image acquisition device 1000 may be a medical image in any format other than the DICOM format, and in this case as well, the image analysis device 2000 will likely be provided to perform conversion operations to any appropriate format.

[0293] For example, the image analysis device 2000 may be provided to remove noise or correct artifacts that may be present in the images acquired from the image acquisition device 1000. For instance, blurring and median filtering techniques can be used to remove noise. By removing noise and correcting artifacts, the image analysis device 2000 can derive more accurate image segmentation results and calculate morphological indicators based on these improved segmentation results, thereby providing objective diagnostic aids for brain-related diseases with high reliability.

[0294] For example, the image analysis device 2000 may be provided to perform an operation to correct the intensity of the image acquired from the image acquisition device 1000. By appropriately correcting the intensity, noise that may be present in the image can be removed, and an image specific to the anatomical structure of the brain to be analyzed can be obtained.

[0295] For example, the image analysis device 2000 may be provided to perform an operation to smooth the image acquired from the image acquisition device 1000. For instance, techniques such as blurring (RPurring) or using a Gaussian filter may be used as methods for smoothing the image.

[0296] For example, the image analysis device 2000 may be provided to adjust the aspect ratio of images acquired from the image acquisition device 1000 or to perform image cropping operations. For instance, the image analysis device 2000 may be implemented to utilize any appropriate cropping technique to crop the image. Alternatively, the image analysis device 2000 may be implemented to utilize any appropriate image resizing technique to adjust the aspect ratio of images, such as on-demand image resizing, lambda image resizing, a resizing method using the CILanczosScaleTransform filter, or a resizing method using the CIFilter.

[0297] As an example, the image analysis device 2000 can perform an operation to convert an image acquired from the image acquisition device 1000 into an image taken under different scan conditions than those under which the original image was taken. For example, the image analysis device 2000 may be provided to perform an operation to convert an image taken under the first scan conditions of the MRI device into an estimated image that would be the same as one taken under the second scan conditions of the MRI device. Such an image conversion operation may be performed using MR conversion technology or an artificial intelligence model. However, it is not limited to this, and the image analysis device 2000 may be provided to perform image conversion taking scan conditions into consideration using any appropriate software or image processing technology.

[0298] For example, the image analysis device 2000 may be implemented to perform preprocessing operations corresponding to the image preprocessing operations performed by the learning device 2200, which will be described later. For instance, if the learning device 2200 has trained a neural network model using a first preprocessing technique, the image analysis device 2000 may be implemented to preprocess the target image using a preprocessing technique corresponding to the first preprocessing technique. This allows for more stable and accurate image segmentation using the neural network model.

[0299] The image alignment-related operations of an image analysis device 2000 according to one embodiment of this application will be described below with reference to Figures 5 to 6.

[0300] Figure 5 shows an example of the image alignment operation of the image analysis device 2000.

[0301] Figure 6 shows an example of the image alignment operation of the image analysis device 2000.

[0302] The image analysis device 2000 may be implemented to perform an operation to align brain images based on data related to the orientation of the brain images contained in the brain images.

[0303] For example, the image analysis device 2000 may be implemented to perform an alignment operation of brain images before performing an image segmentation operation.

[0304] For example, referring to Figure 5, the image acquisition device 1000 can acquire data (i, j, k) related to the direction of the captured image, taking into account the direction of the reference coordinate axis (RAS) of the target object 100. Specifically, the image acquisition device 1000 can acquire data (i, j, k) related to the direction of the captured image, taking into account the coordinate axis information (x, y, z) of the image acquisition device 1000 and the reference coordinate axis information (RAS) of the target object 100.

[0305] Image orientation-related data can be structured as metadata for the image, or it can be transmitted separately to the image analysis device 2000.

[0306] The image analysis device 2000 may be implemented to align images to correspond to the RAS direction (Right-Anterior-Superior direction) of the object 100, based on data (ijk) related to the image direction.

[0307] According to the image alignment operation of the image analysis device 2000 described above, the brain images that form the basis of segmentation can be aligned in a common direction before performing the segmentation operation. This prevents inaccurate segmentation results and ensures the stability of the neural network segmentation operation.

[0308] The image analysis device 2000 may be implemented to perform spatial normalization of brain images.

[0309] Specifically, artificial neural network models used for brain image segmentation and other purposes can be reliably driven for brain images that correspond to the spatial distribution of the training images used as "training data." In other words, if the brain images differ from the spatial distribution of the training images, there is a possibility that the trained artificial neural network will not be able to be driven reliably.

[0310] Therefore, the image analysis device 2000 may be implemented to perform spatial normalization of brain images in order to reduce the spatial uncertainty of brain images.

[0311] As an example, an image analysis device 2000 according to one embodiment of this application may be implemented to perform spatial normalization of brain images based on a brain template. Specifically, the image analysis device 2000 can transform the coordinates of brain images so that the spatial distribution of brain images is optimized for an artificial neural network model by aligning the brain images with a brain template.

[0312] For example, brain templates used to reconcile brain images could include MNI templates, Talairach templates, and so on.

[0313] As described above, the image analysis device 2000 performs an operation to align brain images with brain templates. In this case, "alignment" may not mean an operation to match the internal brain elements of the brain images with the internal brain elements of the brain templates, but simply aligning the spatial position of the brain images.

[0314] According to the spatial normalization operation of the image analysis device 2000 described above, spatial uncertainty in the image can be eliminated, ensuring the stability of segmentation operations using a neural network, and enabling the acquisition of segmentation results with improved accuracy.

[0315] Furthermore, the consistency of the brain image with respect to the brain template allows for the generation of data associated with the transformed coordinates. This data associated with the transformed coordinates can be used to convert the brain image back to its original coordinates after the segmentation described later is complete. In other words, the image analysis device 2000 may be implemented to use the data associated with the transformed coordinates to perform an operation that converts the segmentation results obtained based on the spatial distribution of the brain template to correspond to the original brain image.

[0316] Through the operation of such an image analysis device 2000, the user can be provided with analysis results regarding the brain image relative to its original coordinates.

[0317] For example, when outputting morphological indicators associated with a brain image via an output module, the output can be morphological indicators corrected using correction parameters. Conversely, when outputting visual information of the brain (e.g., info.13 shown in Figure 57, or segmentation results, etc.) via an output module, the output can be an inversely transformed image based on the transformed coordinates described above, i.e., visual information associated with the brain based on the original brain image.

[0318] In the following, with reference to Figures 7 to 17, the operation of the learning device 2200 and the image analysis device 2000 according to one embodiment of this application, related to image segmentation, will be described.

[0319] According to one embodiment of this application, the image segmentation operation may be performed using a learned neural network model. However, the image segmentation operation according to one embodiment of this application may be realized using any suitable method without using a neural network model.

[0320] The following section will primarily explain the process of training a neural network model for image segmentation and then performing image segmentation using the trained neural network model.

[0321] Refer to Figure 7. Figure 7 is a diagram showing the process flow for image segmentation according to one embodiment of the present invention.

[0322] Referring to Figure 7, an image segmentation process according to one embodiment of the present application may include a learning process P1000 for an artificial neural network model for image segmentation, and a segmentation process P2000 for a target image using the learned artificial neural network model.

[0323] At this time, the learning process P1000 can be realized by the learning device 2200 according to one embodiment of this application.

[0324] Furthermore, the segmentation process P2000 can be realized by an image analysis device 2000 according to one embodiment of this application.

[0325] At this time, the parameters of the neural network model acquired by the learning process P1000 implemented in the learning device 2200 can be transmitted to the image analysis device 2000 via any appropriate communication module.

[0326] At this time, the image analysis device 2000 may be implemented to perform segmentation of the target image based on the parameters of the neural network model acquired by the learning process P1000.

[0327] A learning process P1000 according to one embodiment of this application may include a process P1100 for acquiring an image dataset, a process P1200 for training a neural network model, a process P1300 for validating the neural network model, and a process P1400 for acquiring parameters of the neural network model.

[0328] The learning method for a neural network model of the learning device 2200 according to one embodiment of this application will be described below with reference to Figure 8. Figure 8 is a sequence diagram of the learning method for a neural network model of the learning device 2200 according to one embodiment of this application.

[0329] Referring to Figure 8, a learning method for a neural network model of a learning device 2200 according to one embodiment of this application may include a step S1100 for acquiring an image dataset, a step S1200 for screening the image dataset, a step S1300 for preprocessing and aligning the image dataset, a step S1400 for learning and verifying the neural network model, and a step S1500 for acquiring neural network model parameters.

[0330] In step S1100, which involves acquiring an image dataset, the learning device 2200 according to one embodiment of this application can acquire an image dataset from the image acquisition device 1000 or any external device.

[0331] Refer to Figure 9. Figure 9 is an exemplary structural diagram of an image dataset according to one embodiment of this application.

[0332] The image dataset DS acquired by the learning device 2200 may include at least one image data. In other words, the image dataset DS acquired by the learning device 2200 may include at least one image data, such as a first image data ID1, a second image data ID2, and n image data IDn.

[0333] At this time, the image data included in the image dataset DS acquired by the learning device 2200 may include data related to the image and label.

[0334] For example, referring to Figure 9, the first image data ID1 included in the image dataset DS may include the first image I1 and the first label L1, as well as related data.

[0335] Specifically, the first label L1 can be obtained by manually labeling the first image I1 from a clinician capable of diagnosing brain diseases. Alternatively, the first label L1 can be automatically labeled and obtained using any appropriate image segmentation technique.

[0336] The image and label-related data contained in the image data can be used as a basis for training an artificial neural network model in connection with the learning method according to one embodiment of this application, and for validating the artificial neural network model.

[0337] On the other hand, the image data included in the image dataset DS may include data related to scan conditions. In this case, the data related to scan conditions may be data related to magnetic field strength, setting parameters of the imaging device, and / or data related to the manufacturer of the imaging device, as described above. Furthermore, the data related to scan conditions can be structured with metadata for the image data.

[0338] As an example, referring to Figure 9, the first image data ID1 may include data related to the first scan conditions (SC1) associated with the scan conditions under which the first image data ID1 was acquired. For example, if the first image data ID1 was acquired under a magnetic field strength of 3T, information related to the magnetic field strength corresponding to 3T can be structured as metadata for the first image data ID1 and acquired by the learning device 2200 via the dataset acquisition step S1100.

[0339] Data related to scan conditions included in the image data may be considered in order to obtain correction parameters for correcting morphological values ​​or morphological indices corresponding to target elements obtained from the target image, as described later.

[0340] Figure 9 shows only the data included in the first image data ID1, but this is merely an example. Image datasets including the second image data ID2 or the nth image data IDn, etc., can include image, label, and data related to scan conditions.

[0341] Furthermore, the learning device 2200 may be implemented to acquire information related to brain atlases (atlases) that relate to brain structure or brain function for segmentation associated with brain images.

[0342] Specifically, the learning device 2200 can acquire information related to the brain atlas and brain function described above from the image analysis device 2000 or any external device.

[0343] At this time, the learning device 2200 may be implemented to train the artificial neural network model described later, taking into account the brain atlas and related information, or to verify the artificial neural network model.

[0344] In step S1200, which screens the image dataset, the learning device 2200 according to one embodiment of this application may be implemented to screen the image dataset acquired in step S1100, or to select only some of the image data from among the image data included in the image dataset.

[0345] For example, some image data from acquired image datasets may not be suitable for training an artificial neural network model for segmentation. For instance, some image data may contain significant artifacts or noise. Such image data may not be suitable for training an artificial neural network model.

[0346] Therefore, the learning device 2200 may be implemented to screen image data included in the acquired image dataset or to select image data that is effective for training an artificial neural network model.

[0347] In step S1300, which involves preprocessing and aligning the image dataset, the learning device 2200 according to one embodiment of this application may be implemented to perform preprocessing operations to remove noise and artifacts from the images included in the image dataset or to correct the intensity of the images.

[0348] Furthermore, the learning device 2200 according to one embodiment of this application may be implemented to align images based on data related to the orientation of the images, or to align images by aligning them to a brain template and performing spatial normalization.

[0349] In this regard, the preprocessing operation of the image analysis device 2000 described above and the image alignment operation described above in relation to Figures 5 to 6 may be provided to be implemented in the learning device as well. Alternatively, the image preprocessing and alignment operations may be performed in the image analysis device 2000 via data transmission and reception between the learning device 2200 and the image analysis device 2000, and then transmitted to the learning device 2200.

[0350] In the neural network model learning and validation step S1400, the learning device 2200 for image segmentation according to one embodiment of this application can be trained to learn an artificial neural network model for image segmentation.

[0351] Specifically, the artificial neural network model may include an input layer for receiving image data, an output layer for outputting labeling results which are segmentation results, and a hidden layer containing at least one node.

[0352] In this case, the learning device 2200 may be implemented to receive image data included in the acquired image dataset via the input layer and to obtain labeling results for the image data acquired by the neural network model via the output layer.

[0353] For example, the learning device 2200 may be implemented to learn an artificial neural network configured to take first image data ID1 as input and output a first' label (L1') via an output layer. Alternatively, the learning device 2200 can take second image data ID2 as input in an input layer and acquire the second' label (L2') output via an output layer.

[0354] Furthermore, the learning device 2200 according to one embodiment of this application can learn an artificial neural network by taking into account the brain atlas related to the structure of the brain and the brain atlas related to the function of the brain as described above.

[0355] A learning device according to one embodiment can train a neural network model to segment brain images based on a predetermined brain atlas. The brain atlas may include multiple brain regions, including a first brain region and a second brain region. The learning device can use an image in which the first brain region corresponding to the first brain region and the second brain region corresponding to the second brain region are labeled, and use the image as input to train a neural network model to acquire the first and second brain regions.

[0356] For example, the learning device 2200 according to one embodiment of this application may be implemented to train an artificial neural network model for segmentation of image data included in an image dataset, based on the Desikan-Killiany Atlas (Desikan 2006). The Desikan-Killiany Atlas (Desikan 2006) is an atlas used to acquire regions corresponding to multiple brain regions in a cerebral cortex that includes multiple brain regions, including a first brain region and a second brain region.

[0357] At this time, the learning device 2400 can be trained to learn a neural network model for segmenting the cortical regions of the brain by using image data in which the first region corresponding to the first brain region and the second region corresponding to the second brain region are labeled, taking into account the Desikan-Killiany Atlas (Desikan 2006).

[0358] However, the brain atlas mentioned above is merely an example, and any appropriate brain atlas may be considered depending on the purpose of image data segmentation and the area of ​​interest.

[0359] Furthermore, the learning device 2200 according to one embodiment of this application may be implemented to train a neural network model based on the scanning conditions under which the image data was captured.

[0360] For example, the learning device 2200 may be implemented to utilize a first neural network model to segment multiple regions from a first image acquired under a first scan condition. Conversely, the learning device 2200 may be implemented to utilize a second neural network model to segment multiple regions from a second image acquired under a second scan condition. In other words, the learning device 2200 can be trained to train different neural network models based on the scan conditions under which the images were captured.

[0361] Furthermore, in the segmentation process P2000, the image analysis device 2000 according to one embodiment of this application may be implemented such that target images acquired under first scan conditions are segmented using a learned first neural network model, and target images acquired under second scan conditions are segmented using a learned second neural network model.

[0362] As a result, the image analysis device 2000 according to one embodiment of this application can perform target image segmentation using an optimal neural network model for each scanning condition, thereby enabling the acquisition of multiple regions more stably and accurately.

[0363] In the following, with reference to Figures 10 to 11, examples of artificial neural network models that can utilize the learning device 2200 according to one embodiment of this application will be described.

[0364] Figure 10 shows an example of an artificial neural network model in which the learning device 2200 according to one embodiment of this application can be used.

[0365] Figure 11 shows another example of an artificial neural network model in which the learning device 2200 according to one embodiment of this application can be used.

[0366] Referring to Figure 10, the learning device 2200 according to one embodiment of this application can utilize U-net as an artificial neural network for image segmentation.

[0367] The U-net used for image segmentation may be configured with an architecture that includes a contraction path and an expansion path.

[0368] Specifically, the U-net condensation path may be configured such that two convolutions and max pooling operations are performed sequentially. In this case, the U-net condensation path can extract characteristics related to the image.

[0369] However, since the size of the characteristic map also decreases in the contraction path, the U-net may be configured to include an expansion path to restore the size of the characteristic map.

[0370] The U-net extension path may be configured so that up-convolution and double convolution are performed sequentially. In this case, the U-net extension path can extract the size of the image and the characteristic map.

[0371] Furthermore, U-net's architecture is configured to concatenate characteristic maps at the same level, allowing it to provide characteristic-related location information from contraction paths to expansion paths.

[0372] At this time, based on the difference between the labels of the input image and the labels of the output segmentation map, the parameters or weights of at least one node in the layer containing the U-net can be adjusted so that the difference between the labels of the input image and the labels of the output segmentation map is minimized.

[0373] Furthermore, referring to Figure 11, the learning device 2200 according to one embodiment of this application can utilize U-net++ in an artificial neural network for image segmentation. U-net++ differs from U-net in that it is an artificial neural network model that uses the high-density block idea of ​​DenseNet to improve the performance of U-net, and has convolutional layers in the skip paths to connect the semantic gap between the characteristic maps of encoders and decoders, and has dense skip connections in the skip paths to improve gradient flow.

[0374] Specifically, the learning device 2200 may be implemented to input an input image to the input layer of the U-net++ neural network model and to acquire label information output via the output layer. In this case, the learning device 2200 can adjust the parameters or weights of at least one node in the hidden layer containing Unet++ based on the difference between the label information contained in the input image and the label information output from the neural network model.

[0375] Specifically, the learning device 2200 is implemented to repeatedly perform the operation of adjusting the parameters or weights of at least one node as described above, and can obtain the parameters or weights of the node that minimizes the difference between the label information contained in the input image and the label information output from the neural network model.

[0376] As described above, the learning device 2200 according to one embodiment of this application can perform an operation to train an artificial neural network model based on the outputted label results.

[0377] Specifically, in step S1400, which trains the artificial neural network model, data related to the labels contained in the image data obtained from step S1100, which acquires the image dataset, can be obtained.

[0378] In this case, the learning device 2200 may be implemented to train an artificial neural network model based on image data and label data output via the output layer of the neural network model.

[0379] More specifically, the learning device 2200 may be implemented to train the neural network model by adjusting the weights and parameters of at least one node in the hidden layer of the neural network model based on the difference between the label data contained in the image data and the label data output through the output layer of the neural network model.

[0380] For example, the learning device 2200 can input the first image data ID1 into the input layer of the artificial neural network and obtain label data corresponding to the first A label L1A. At this time, the learning device can train a neural network model based on the label data corresponding to the first label L1 and the label data related to the first A label L1A contained in the first image data ID1. For example, the learning device 2200 may be implemented to train the neural network model by adjusting the weights and parameters of at least one node included in the hidden layer of the neural network model based on the difference between the first label L1 and the first A label L1A.

[0381] As another example, the learning device 2200 can input the i-th image data IDi into the input layer of the artificial neural network to obtain label data corresponding to the iA label LiA. In this case, the learning device can train a neural network model based on the label data corresponding to the i-th label Li and the label data related to the iA label LiA contained in the i-th image data IDi. For example, the learning device 2200 may be implemented to train the neural network model by adjusting the weights and parameters of at least one node included in the hidden layer of the neural network model based on the difference between the i-th label Li and the iA label LiA, where i can be any number.

[0382] In step S1400, which verifies the artificial neural network model, the learning device 2200 according to one embodiment of this application can verify the artificial neural network model.

[0383] As an example, the learning device 2200 according to one embodiment of this application can acquire label data output via a learned neural network model based on at least one image data contained in the image dataset DS. At this time, the learning device 2200 can verify the learned neural network model based on the label data associated with at least one image data and the label data output via the learned neural network model.

[0384] For example, the learning device 2200 can verify whether the parameters or weights of the nodes in the hidden layer of the learned neural network model are appropriate by comparing the similarity between at least one image data and associated label data and the label data output via the learned neural network model.

[0385] In step S1500, which involves acquiring an artificial neural network model, the learning device 2200 according to one embodiment of this application repeatedly performs operations to train the artificial neural network model on image data included in the image dataset and to validate the artificial neural network model, thereby acquiring a neural network model that includes at least one node having weights and parameters that minimize the difference between data related to labels included in the image data and data related to labels output by the artificial neural network.

[0386] The acquired node parameters and weights can be used in the artificial neural network model for image segmentation in the P2000 segmentation process.

[0387] As described above, the focus has been on segmentation using an artificial neural network, but the image analysis device 2000 disclosed in this application can utilize various image segmentation algorithms, including image segmentation using an artificial neural network.

[0388] For example, an image segmentation algorithm may be provided as a machine learning model. A typical example of a machine learning model is an artificial neural network. Specifically, a typical example of an artificial neural network is the deep learning series, which includes an input layer that receives data input, an output layer that outputs results, and a hidden layer that processes data between the input and output layers. Detailed examples of artificial neural networks include convolutional neural networks, recurrent neural networks, deep neural networks, and generative adversarial networks. In this specification, the term artificial neural network should be interpreted in a comprehensive sense, including the artificial neural networks described above, various other forms of artificial neural networks, and combinations thereof, and is not necessarily limited to the deep learning series.

[0389] Furthermore, machine learning models are not necessarily limited to artificial neural network (MS) models; they may also include the nearest neighbor algorithm (KNN), random forest (RandomForest), support vector machines (SVM), principal component analysis (PCA), and others. The techniques mentioned above can be combined in ensemble forms or in various other ways. On the other hand, in examples that primarily refer to MS, unless otherwise specified, MS may be replaced by other machine learning models.

[0390] Furthermore, in this specification, image segmentation algorithms are not necessarily limited to machine learning models. In other words, image segmentation algorithms may include various judgment and decision algorithms that are not machine learning models.

[0391] Therefore, in this specification, the term "image segmentation algorithm" should be understood in a comprehensive sense, encompassing all forms of algorithms that perform segmentation using image data.

[0392] Referring again to Figure 7, the segmentation process P2000 according to one embodiment of the present application may include a data acquisition process P2100 and a segmentation process P2200 that utilizes a learned neural network model.

[0393] The segmentation process P2000 may be implemented by an image analysis device 2000 according to one embodiment of this application.

[0394] The following describes the image segmentation operation of the image analysis device 2000 according to one embodiment of this application, utilizing a neural network model, with reference to Figure 12. Figure 12 is a sequence diagram of the image segmentation method using a neural network model of the image analysis device 2000 according to one embodiment of this application.

[0395] Referring to Figure 12, an image segmentation method using a neural network model of an image analysis device 2000 according to one embodiment of this application may include a step S2100 for acquiring target image data, a step S2200 for acquiring segmentation information using the learned neural network, and a step S2300 for outputting segmentation information.

[0396] Specifically, in step S2000, which involves acquiring target image data, the image analysis device 2000 can acquire the target image from the image acquisition device 1000. Furthermore, the image analysis device 2000 can acquire target object information related to the target image, or information related to the scan conditions under which the target image was captured, from the image acquisition device 1000 or any external device.

[0397] At this time, information about the target object associated with the target image, or information related to the scan conditions under which the target image was captured, may be considered when acquiring the correction parameters described later. This will be explained in detail in relation to Figures 58 to 67.

[0398] Refer to Figure 13. Figure 13 is an exemplary structural diagram of a target image according to one embodiment of this application.

[0399] As an example, the target image data acquired by the image analysis device 2000 according to one embodiment of this application may include information about the target image TI. For example, the information about the target image TI may include information related to pixel coordinates, intensity, color, etc.

[0400] As another example, the target image data may include target object information TO. For example, the information for target object information TO may be information about personal information of the subject (e.g., a patient undergoing brain disease examination) associated with the target image TI. For example, the information for target object information TO may be information related to the name, age, gender, etc., of the subject (e.g., a patient undergoing brain disease examination). In this case, the image analysis device 2000 can obtain information for target object information TO from any external device. Alternatively, the image analysis device 2000 can obtain information for target object information TO by recognizing structured metadata for the target image data.

[0401] As another example, the target image data may include information about the target scan conditions TSC. For example, the information about the target scan conditions TSC may be related to the scan conditions under which the target image TI was captured. For example, the information about the target scan conditions TSC may be related to the magnetic field strength under which the target image TI was captured, the setting parameters of the imaging device under which the target image TI was captured, or the manufacturer of the imaging device under which the target image TI was captured. In this case, the image analyzer 2000 can obtain information about the target scan conditions TSC from any external device. Alternatively, the image analyzer 2000 can obtain information about the target scan conditions TSC by obtaining structured metadata for the target image data.

[0402] Referring again to Figure 7, the image analysis device 2000 may be implemented to input the target image data acquired in the data acquisition process P2100 into the input layer of the learned neural network model.

[0403] In this case, the image analysis device 2000 may be implemented to utilize the node weights and / or node parameters of the artificial neural network model acquired based on the learning process P1000 implemented by the learning device 2200 described above in the artificial neural network model for segmenting the target image data.

[0404] Referring again to Figure 12, in step S2200, which involves acquiring segmentation information using a learned neural network model, the learned neural network model, based on the node weights and / or node parameters acquired from the learning device 2200, may be provided to receive target image data as input via an input layer and, as a result of segmenting the target image TI, output the result of labeling the target image TI via an output layer. At this time, the image analysis device 2000 can acquire segmentation information related to the labeling of the target image via the output layer of the learned neural network model.

[0405] In this case, the results output via the output layer of the artificial neural network model may include multiple target regions obtained from the target image of the target image data.

[0406] The output of the neural network model may be in the form of labeling corresponding to multiple target regions obtained from the target image. For example, the output of the neural network model may be in the form of label data including a first label defining a first region and a second label defining a second region obtained from the target image.

[0407] In this case, the output via the neural network model's output layer may be in a form where the first color is overlaid on the first region of the target image based on the first label, and the second color is overlaid on the second region of the target image. This makes it easier to distinguish between the first and second regions. However, the above is merely an example, and the output result can be configured in any form necessary to distinguish between the first and second regions.

[0408] Furthermore, the segmentation information output via the output layer of the neural network model may be in a form in which the target image is divided into multiple regions based on a predetermined brain atlas. For example, as described above, the learning device can learn the neural network model to divide the learning image into a first region corresponding to the first brain region and a second region corresponding to the second brain region based on a predetermined brain atlas. In this case, since the neural network model learned based on a predetermined brain atlas is used, the segmentation information output via the output layer of the neural network model may be output in a form in which the target image is divided into multiple regions, including the first region corresponding to the first brain region and the second brain region, as well as the second region.

[0409] The image analysis device 2000 can perform the operation of calculating morphological values ​​corresponding to a specific region based on the segmentation information output via the output layer.

[0410] For example, the image analysis device 2000 may be implemented to calculate morphological numerical values ​​representing morphological characteristics associated with a first brain region based on segmentation information related to a first region output via an output layer. For example, morphological characteristics can relate to volume, thickness, length, or shape.

[0411] In step S2300, which outputs segmentation information, the image analysis device 2000 may be configured to overlay visual graphics onto multiple brain regions based on the segmentation information output via the output layer and display them to the user via the output module 2650 of the output device 2600 or the output module 2050 of the image analysis device 2000.

[0412] Refer to Figure 14. Figure 14 is an example of an image obtained based on segmentation information acquired by the segmentation process of an image analysis device 2000 according to one embodiment of this application.

[0413] Specifically, the diagram shown at the top of Figure 14 is an example of an image output based on segmentation results for images acquired from T1-MRI. The diagram shown at the bottom of Figure 14 is an example of an image output based on segmentation results for images acquired from T2-Flair MRI.

[0414] For example, the image analysis device 2000 according to one embodiment of this application can acquire segmentation information corresponding to the frontal lobe, temporal lobe, parietal lobe, occipital lobe, lateral ventricle, amygdala, hippocampus, etc., by a segmentation process applied to images acquired from T1-MRI.

[0415] For example, the image analysis device 2000 according to one embodiment of this application can acquire segmentation information corresponding to white matter, gray matter, ventricles, WMH (white matter hyperintensity) regions, etc., by a segmentation process applied to images acquired from T2-Flair MRI.

[0416] According to the image segmentation operation of the image analysis device 2000 described above, the user can visually confirm the segmentation results and easily verify them. Furthermore, it provides the advantageous effect of improving the user's understanding of auxiliary indicators for diagnosing brain diseases.

[0417] On the other hand, the artificial neural network model used in Figure 7 may be implemented using at least one artificial neural network model.

[0418] The following describes a flowchart of the image segmentation process according to one embodiment of this application that utilizes at least one artificial neural network model, with reference to Figures 15 to 17. Specifically, the characteristic features of using multiple artificial neural network models will be described in relation to the process P1200 for training the artificial neural network model in Figure 7. The content described in relation to Figures 7 to 14 can also be applied by analogy to the embodiments described later in relation to Figures 15 to 17.

[0419] Figure 15 shows the process flow for image segmentation according to this embodiment.

[0420] As an example, referring to Figure 15, the neural network model learning process P1200 of the learning process P1000 in Figure 7 may consist of a process P1210 for learning a first neural network model and a process P1220 for learning a second neural network model.

[0421] For example, the first neural network model may be trained to acquire the first region corresponding to the first brain region and the second region corresponding to the second brain region. In this case, the second neural network model may be trained to acquire the third and fourth regions included in the first region that can be acquired from the first neural network model. In other words, the first neural network model may be trained to acquire regions corresponding to the macroscopic structure of the brain (e.g., the cranial region, the cerebrospinal fluid (CSF) region, the cortical region, and the medullary region), and the second neural network model may be trained to acquire regions corresponding to the detailed structure of the brain (e.g., regions corresponding to brain elements located within the cortex, and regions corresponding to brain elements located within the medulla).

[0422] As another example, the first neural network model may be trained to acquire regions corresponding to brain regions corresponding to the first brain atlas. Conversely, the second neural network model may be trained to acquire regions corresponding to brain regions corresponding to the second brain atlas.

[0423] As another example, the first neural network model may be trained to acquire regions corresponding to the macroscopic structure of the brain (e.g., cranial regions, cerebrospinal fluid (CSF) regions, cortical regions, and medullary regions). Conversely, the second neural network model may be trained to acquire regions corresponding to brain regions corresponding to brain atlases. For example, the second neural network model may be trained to acquire regions corresponding to multiple brain regions based on the Desikan-Killiany Atlas (Desikan 2006), and the second neural network model trained on the Desikan-Killiany Atlas (Desikan 2006) may be trained to acquire regions including the frontal lobe, temporal lobe, parietal lobe, occipital lobe, lateral ventricles, amygdala, and hippocampus.

[0424] The process P1210 for training the first neural network model and the process P1220 for training the second neural network model may be performed independently.

[0425] Specifically, the datasets used in process P1210 for training the first neural network model and process P1220 for training the second neural network model can be independent.

[0426] Refer to Figure 16. Figure 16 is an illustrative structural diagram of an image dataset according to one embodiment of this application. Specifically, the image dataset shown on the left side of Figure 16 may be an image dataset for training a first neural network model. Conversely, the image dataset shown on the right side of Figure 16 may be an image dataset for training a second neural network model.

[0427] For example, an image dataset DS for training a first neural network model may include a first image I1 and a first a label L1A associated with a first region, along with related information. Conversely, an image dataset DS for training a second neural network model may include a first image I1 and a first b label L1b associated with a second region, along with related information.

[0428] At this time, the learning device can use the first neural network model to input the first image data ID1 into the input layer of the artificial neural network and obtain output label data corresponding to the first a label L1a' associated with the first region. At this time, the learning device can train the artificial neural network model based on the label data corresponding to the label L1A associated with the first region contained in the first image data ID1 and the output label data associated with the first a' label L1a' associated with the first region. For example, the learning device 2200 may be implemented to adjust the weights and parameters of at least one node included in the hidden layer of the neural network model based on the difference between the first a label L1A and the first a' label L1a'. Furthermore, by repeatedly executing the process of training the first neural network model described above, parameters related to the first neural network model associated with the first region can be obtained.

[0429] Conversely, the learning device can use the second neural network model to input the first image data into the input layer of the artificial neural network and obtain output label data corresponding to the first b' label L1b' associated with the second region. At this time, the learning device can train the artificial neural network model based on the label data corresponding to the first b label L1b associated with the second region and the output label data associated with the first b label L1b' associated with the second region, which are contained in the first image data ID1. For example, the learning device 2200 may be implemented to adjust the weights and parameters of at least one node contained in the hidden layer of the neural network model based on the difference between the first b label L1b and the first b' label L1b'. Furthermore, by repeatedly executing the process of training the second neural network model described above, parameters related to the second neural network model associated with the second region can be obtained.

[0430] Furthermore, parameters related to the first neural network model acquired in the learning process P1000 may be used in the first neural network model for image segmentation corresponding to the first region of the segmentation process P2000 (P2210), and parameters related to the second neural network model acquired in the learning process P1000 may be used in the second neural network model for image segmentation corresponding to the second region of the segmentation process P2000 (P2220).

[0431] However, as shown in Figure 16, the 1a label included in the image data related to the first neural network model is a label related to the first region, and the 1b label included in the image data related to the second neural network model is a label related to the second region. This is merely an example and is not limited to this. For example, the image data of the first and second neural network models may include label information related to the first and second regions, and when training the first neural network model, only the label information related to the first region may be used, and when training the second neural network model, only the label information related to the second region may be used.

[0432] Furthermore, although the first and second neural network models were described as independent entities as described above, this is not the only explanation; the first and second neural network models can share at least some layers. In other words, the first and second neural network models can contain at least one common layer.

[0433] On the other hand, the process of training the first neural network model and the process of training the second neural network model may be independent, or they may be trained in a manner related to each other. Here, training them in a manner related to each other can mean using the output data from one of the two neural network models as input data for the other neural network model, and including all forms in which any data generated by one of the two neural network models is used by the other neural network model.

[0434] Refer to Figure 17. Figure 17 is a diagram showing the process flow for image segmentation according to an embodiment of this application.

[0435] As one embodiment, the process P1200 for training a neural network model of the learning process P1000 according to one embodiment of this application may include a process for training a first neural network model (P1211) and a process for training a second neural network model (P1221), and may be implemented such that data related to the results output from the first neural network model is input to the second neural network model as input data.

[0436] For example, the learning device 2200 can be configured to receive input of label data related to the first region in the second neural network model. In this case, the label data related to the first region can be obtained by manually labeling the first brain region corresponding to the image data, or by using any automatic labeling software.

[0437] For example, the label data associated with the first region may be data that has been manually labeled by a clinician for the image data input into the first neural network model.

[0438] In another example, the first neural network model can be trained to output label data associated with the first region, and the label data associated with the first region that is input to the second neural network model may be the label data output from the first neural network model.

[0439] At this time, the second neural network model can be trained to output label data related to the second region based on label data and image data related to the first region.

[0440] Therefore, the learning device 2200 according to one embodiment of this application may be implemented to acquire label data associated with a first region via a first neural network model and to acquire label data associated with a second region via a second neural network model. Thus, the learning device 2200 can be trained to acquire label data associated with the first and second regions.

[0441] Furthermore, the learning device 2200 may be provided to adjust the weights and parameters of at least one node of the first neural network model or at least one node of the second neural network model based on the difference between the label data associated with labels associated with the first and second regions and the label data associated with the first and second regions of the image dataset.

[0442] Furthermore, parameters related to the first and second neural network models acquired in the learning process P1000 may be used for the first and second neural network models for image segmentation, corresponding to the first and second regions of the segmentation process P2000.

[0443] In this case, the first or second region may be a region associated with the anatomical structure of the brain, or it may be one of the regions demarcated based on the brain atlas mentioned above. Furthermore, the first or second region may be a region that shows a meaningful association with brain diseases.

[0444] However, the learning process of the neural network model described above is merely illustrative. It would be possible to learn image segmentation behavior by utilizing various combinations and connectivity relationships of at least one or more neural network models having any appropriate form, type, and parameters, validate the neural network model, and obtain the parameters of the neural network model.

[0445] On the other hand, an image analysis device 2000 according to one embodiment of this application may be implemented to update or renew a learned artificial neural network model in the segmentation process P2000.

[0446] As an example, the image analysis device 2000 may be implemented to allow the segmentation information obtained by segmenting a target image using a segmentation process P2200 that utilizes a learned neural network model to be modified manually or using any software. In this case, the image analysis device 2000 may be implemented to update or update the artificial neural network model by modifying the weights of at least one node or the parameters of at least one node of the learned neural network model based on the modified segmentation information.

[0447] An image analysis device according to one embodiment can perform a function to determine the quality of an image. The image analysis device has quality criteria for determining the quality of an image and can determine whether or not the input image satisfies the quality criteria.

[0448] More specifically, image quality assessment can mean determining whether medical images acquired by an image acquisition device possess a certain level of quality or higher. In other words, image quality assessment can mean determining whether it is possible to obtain medical information with a certain level of reliability or higher from medical images acquired by an image acquisition device.

[0449] Image quality determination may be performed in conjunction with the detailed operations that constitute the image analysis operation. For example, the image quality determination operation may be performed in conjunction with the image acquisition operation, image preprocessing operation, image segmentation operation, and / or medical information output operation described herein.

[0450] The image quality judgment operation may be performed before or after the execution of at least one detailed operation. The image quality judgment operation may be performed based on information obtained from the results of the detailed operation. The image quality judgment operation may be performed to determine whether the criteria for performing the detailed operation are met.

[0451] Image quality assessment may be performed based on image data. Image quality assessment may be performed based on raw image data. Image quality assessment may be performed based on preprocessed medical images. As another example, image quality assessment may be performed based on the results of segmentation of medical images. As yet another example, image quality assessment may be performed based on the results of analysis of medical images.

[0452] Image quality assessment may be performed based on non-image data associated with image data. Image quality assessment may be performed based on at least one of the following: metadata information related to medical images, artifact information contained in medical images, or region of interest information obtained through medical image segmentation.

[0453] According to one embodiment, image quality determination may be performed separately for each corresponding detailed operation.

[0454] For example, the first quality judgment may be performed based on raw data before preprocessing operations. In this case, preprocessing operations may be performed on images that satisfy the first quality criterion. Alternatively, for example, the second quality judgment may be performed based on raw data or preprocessed images before segmentation operations. In this case, segmentation operations may be performed on images that satisfy the second quality criterion. Furthermore, for example, the third quality judgment may be performed based on raw data, preprocessed images, or image segmentation results before image analysis. Image analysis may be performed on images that satisfy the third quality criterion. The first to third quality criteria can each be different.

[0455] Information obtained through image quality assessment can be output. This information may include the basis for the image quality assessment. This basis may include whether or not defects exist in the image, image format information, and image acquisition conditions. The information obtained through image quality assessment can be used to generate or provide information related to whether or not to perform subsequent actions.

[0456] Specific examples of how image quality assessment is performed will be described later.

[0457] An image analysis device 2000 according to one embodiment of this application may be provided to perform the operation of calculating brain morphological indicators of a target body based on the segmentation results of a target image.

[0458] Specifically, the image analysis device 2000 can acquire the internal region of the skull and the region corresponding to the target element based on the segmentation results. In addition, the image analysis device 2000 can perform an operation to correct the boundary corresponding to the internal region of the skull in order to calculate the morphological index of the target element.

[0459] Furthermore, the image analysis device 2000 can additionally perform actions to align brain images in order to correct the boundaries of the internal regions of the skull.

[0460] The operation by which the image analysis device 2000 according to one embodiment of this application calculates brain morphological indicators of a target body will be described in detail later in relation to Figures 48 to 57.

[0461] An image analysis device 2000 according to one embodiment of this application may be provided to perform an operation to correct morphological values ​​or morphological indicators of the brain calculated based on the segmentation results of a target image.

[0462] Specifically, the image analysis device 2000 can acquire the region corresponding to the target element and the pixel or voxel data corresponding to the target element based on the segmentation results. At this time, the image analysis device 2000 can acquire the morphological values ​​of the target element based on the pixel or voxel data corresponding to the target element.

[0463] Furthermore, the image analysis device 2000 can perform an operation to correct the morphological values ​​of target elements by taking into account the scan conditions under which the target image is acquired and the position of target elements within the target image, in order to output more accurate morphological indicators.

[0464] At this time, the image analysis device 2000 may be provided to obtain correction parameters from the correction parameter acquisition device 2400, taking into account the scan conditions and the position of the target element, in order to correct the morphological values ​​of the target element, and to perform an operation to output a morphological index of the target element based on the correction parameters and the morphological values ​​of the target element.

[0465] The operation of the image analysis device 2000 according to one embodiment of this application to correct the morphological values ​​or morphological indicators of the brain will be described in detail later in relation to Figures 58 to 67.

[0466] An image output device according to one embodiment can provide the user with diagnostic support information based on various medical information acquired through image analysis. Here, the diagnostic support information may include information acquired through processing of various medical information obtained from medical images. For example, the diagnostic support information may include information acquired by processing medical information, such as diagnostic information, analysis information, and prescription information based on medical information.

[0467] The diagnostic support information provision operation may be performed together with the detailed operations that constitute the image output operation. For example, the diagnostic support information provision operation may be performed together with the image acquisition operation, the operation to acquire medical information from the image, the operation to acquire diagnostic support information from the acquired information, the operation to output the diagnostic support information, and / or the operation to provide comments based on the diagnostic support information, as described herein. Detailed explanations of each of the operations described above will be given later.

[0468] Image output devices according to other embodiments can selectively provide the user with necessary indicator information from among the various medical information acquired through image analysis. Here, selective information provision may include selectively providing only the medical information necessary for the user from among the various medical information that can be acquired through the image analysis device.

[0469] Selective information provision operations may be performed in conjunction with the detailed operations that constitute the image output operation. For example, selective information provision operations may be performed in conjunction with the image acquisition operation, the operation of acquiring medical information from the image, the operation of acquiring selective information from the acquired information, the operation of outputting the selective information, and / or the operation of providing comments based on the selective information, as described herein. Detailed explanations of each of the operations described above will be given later.

[0470] The configuration and operation of the image analysis apparatus 2000 according to one embodiment of this application have been described above. The image analysis method in this embodiment will be described in more detail below.

[0471] In the following description, the image analysis method according to one embodiment of this application will be described as being performed by the image analysis device 2000, learning device 2200, correction parameter acquisition device 2400, or output device 2600 described above. However, this is merely for the convenience of explanation, and the image analysis method according to one embodiment of this application is not limited to the image analysis device 2000, learning device 2200, correction parameter acquisition device 2400, or output device 2600 described above. In other words, the image analysis method described later is not necessarily performed only by the image analysis device 2000, learning device 2200, correction parameter acquisition device 2400, or output device 2600, and can also be performed by other systems or devices having similar functions to the image analysis device 2000, learning device 2200, correction parameter acquisition device 2400, or output device 2600 described above.

[0472] Various information is obtained from medical images to assess a patient's health status, but in order to obtain accurate information, the analysis must be based on medical images that meet certain quality standards and requirements.

[0473] More specifically, medical images acquired by an image acquisition device may have qualities (or properties) that make them unsuitable for image analysis. For example, images may have formal or substantive defects that make them unsuitable for medical data acquisition. For instance, medical images may not meet a certain standard of quality due to various defects in the image caused by issues such as brightness, resolution, the area being photographed, the direction of photography, the angle of photography, or other problems that may occur during photography. Alternatively, the image format, such as the format and size of the image file, may not meet the necessary requirements.

[0474] When analysis is conducted based on medical images that do not meet a certain standard of quality, the analysis results may also lack a certain level of reliability. Therefore, in order to derive analysis results with a higher level of reliability, an image quality assessment must be performed to determine whether the images acquired by the image acquisition device meet a certain standard of quality.

[0475] Traditionally, inspectors had to manually judge each medical image acquired via imaging equipment to determine if there were any defects. However, this approach had limitations, as the results of the judgment were inconsistent due to variations in the inspector's perspective, experience, and physical condition, making it difficult to achieve a consistently high level of quality assessment.

[0476] According to one embodiment, artificial intelligence can be used to overcome the aforementioned limitations and improve the quality judgment of captured medical images. In other words, the image analysis device according to one embodiment can perform high-level medical image quality judgment by utilizing artificial intelligence.

[0477] Figure 18 is a diagram illustrating the image quality determination process according to one embodiment.

[0478] Referring to Figure 18, an image quality determination process according to one embodiment may include a medical image acquisition step S3000, a medical image quality determination step S3300, and a medical image quality-related information output step S3500. In this case, the medical image quality-related information output step S3500 may be omitted. On the other hand, the image quality determination process may further include a step of preprocessing the medical image and / or a step of segmenting the medical image.

[0479] Figure 19 is a diagram illustrating the image analysis device 3000 that performs the image quality judgment process.

[0480] Referring to Figure 19, the image analysis device 3000 can perform an image quality determination process. The image analysis device 3000 may include one or more modules for performing image quality determination. For example, the image analysis device 3000 may include at least one of the following: a preprocessing execution module 3300, an image segmentation module 3500, or an image quality determination module 3700, or an image quality-related information output module 3900.

[0481] The following sections will describe each step of the image quality assessment process with more specific examples. The image quality assessment process described below may be performed by the image analysis apparatus 3000 mentioned above or by any apparatus or system described throughout this specification.

[0482] An image acquisition device according to one embodiment can acquire medical images. Exemplarily, medical images include, but are not limited to, images from CT (Computed Tomography), MRI (Magnetic Resonance Maging), and X-ray. In this case, MRI images can include various types of images that can be acquired by an MRI scanner, such as T1-weighted images, T2-weighted images, and FLAIR images. Furthermore, MRI images can include images acquired in various planes acquired by an MRI scanner, such as axial, sagittal, or coronal planes.

[0483] An image acquisition device according to one embodiment can acquire medical images in which at least a portion is not completely captured. For example, the image acquisition device can acquire medical images in which at least a portion of the target area is missing or improperly captured.

[0484] For example, an image acquisition device may acquire medical images that are of a quality below a certain level. It may also acquire medical images from which it is unlikely to extract medical information with a reliability above a certain level. Another example is an image acquisition device that may acquire medical images with an abnormal file structure or in which patient information is missing. Yet another example is an image acquisition device that may acquire medical images containing at least one type of noise. Here, noise refers to various types of defects or conditions that affect the acquisition of information based on the image.

[0485] An image quality determination process according to one embodiment may include a step of preprocessing medical images. For example, an image quality determination process according to one embodiment may include a step of preprocessing medical images acquired by an image acquisition device so that they are suitable for performing an image quality determination.

[0486] For example, the image preprocessing step may include performing various preprocessing steps, such as correcting the brightness, size, aspect ratio, orientation, or resolution of the image, so that artifacts contained within the medical image can be detected more easily. As another example, the image preprocessing step may include performing various preprocessing steps, such as correcting the brightness, size, aspect ratio, orientation, or resolution of the image, so that information about anatomical structures contained within the medical image can be obtained more easily.

[0487] In addition, the image preprocessing step may include performing various preprocessing steps for image quality assessment. Details regarding this have been mentioned above, so any redundant information will be omitted.

[0488] An image quality determination process according to one embodiment may include a step of segmenting the medical image. For example, an image quality determination process according to one embodiment may include a step of segmenting the medical image acquired by an image acquisition device for image quality determination.

[0489] Exemplary, the image segmentation step may include performing medical image segmentation to obtain artifacts and associated information contained within a medical image. More specifically, the image segmentation step may include performing segmentation of artifacts and corresponding regions contained within a medical image.

[0490] Segmentation of regions corresponding to artifacts may be performed using a neural network model trained to acquire regions corresponding to artifacts contained in medical images. Segmentation of artifact regions may be performed using a neural network model trained on training data containing one or more medical images in which artifact regions are labeled.

[0491] As another example, the image segmentation step may include segmenting a medical image so that an image quality assessment is made based on at least some of the regions of the human body contained within the medical image. More specifically, the image segmentation step may include segmenting a medical image to obtain information about anatomical or functional structures of the human body that can form the basis for an image quality assessment. The image segmentation step may include segmenting regions of the human body contained within the medical image that correspond to those structures. The image segmentation step may include obtaining regions that correspond to structures used in an image quality assessment.

[0492] Segmentation of regions corresponding to human body structures may be performed using a neural network model. Segmentation of regions corresponding to structures may be performed using a neural network model that has been trained to segment regions contained in medical images. The image segmentation step may include performing segmentation of a medical image using a pre-trained neural network model to obtain at least one segmented region. In this case, each of the at least one segmented region may correspond to another anatomical or functional structure. Furthermore, at least one segmented region may include a region corresponding to a human body structure used for quality assessment. Segmentation of regions corresponding to anatomical or functional structures can be similarly applied to the content related to image segmentation described throughout this specification.

[0493] Figure 20 is a diagram illustrating the image quality judgment module 3700.

[0494] Referring to Figure 20, the image quality determination module 3700 may include at least one of the first image quality determination unit 3710, the second image quality determination unit 3730, the third image quality determination unit 3750, or the fourth image quality determination unit 3770.

[0495] For example, the first image quality determination unit 3710 can determine the image quality based on metadata information, the second image quality determination unit 3730 can determine the image quality based on noise information, the third image quality determination unit 3750 can determine the image quality based on anatomically segmented segmentation information, and the fourth image quality determination unit 3770 can determine the image quality based on composite information, such as the relationship between noise information and anatomically segmented segmentation information. Further details regarding the first to fourth image quality determination units 3710 and 3770 will be described later.

[0496] The neural network model that performs image quality determination according to one embodiment may be learned and executed differently depending on the type of medical image acquired by the image acquisition device. For example, if the type of medical image acquired by the image acquisition device is a CT image, the image quality determination model of the image analysis device 3000 may be a model that is learned and executed based on the CT image. As another example, if the type of medical image acquired by the image acquisition device is an MRI, the image quality determination model of the image analysis device 3000 may be a model that is learned and executed based on the MRI image.

[0497] In one embodiment, the first image quality determination unit 3710 may be executed based on metadata information to determine whether or not the image analysis is performed successfully. In this case, the metadata information may include information about the medical image acquired from the image acquisition device. More specifically, the metadata information may include at least one of the following: file structure information of the medical image or patient information entered into the medical image.

[0498] Figure 21 is a diagram illustrating the first image quality judgment process.

[0499] Referring to Figure 21, the first image quality determination process may include a medical image acquisition step S3711, a non-image information acquisition step S3713, a non-image information normality determination step S3715, and a non-image related information output step S3717.

[0500] The non-image information acquisition step S3713 may include acquiring medical image information from a medical image. The non-image information acquisition step S3713 may include acquiring non-image information from a medical image. Here, the non-image information may include file structure information of the medical image or patient information entered into the medical image.

[0501] In this case, the file structure information of the medical image may include, but is not limited to, information regarding the file format, style, or size of the medical image. Furthermore, patient information may include, but is not limited to, information related to the patient's personal information such as name and age, information related to the time when the medical image of the patient was taken, or information regarding the patient's health condition.

[0502] The non-image information normality determination step S3715 may include performing an image quality determination based on non-image information extracted from the medical image. The non-image information normality determination step S3715 may also include performing an image quality determination based on at least one of file structure information or patient information from the medical image.

[0503] According to one embodiment, the non-image information normality determination step S3715 can determine whether the file structure of the medical image is abnormal. Here, the file structure of the medical image may mean, but is not limited to, the file format of the medical image, the format of the medical image, or the size of the file.

[0504] For example, since the file format or format of the medical image must be a file format or format suitable for image analysis to be performed by the image analysis device 3000, the non-image information normality determination step S3715 may include determining whether the file format or format of the medical image is a file format or format suitable for image analysis to be performed by the image analysis device 3000. As a result, the non-image information normality determination step S3715 may include obtaining information on whether the image analysis will be performed successfully based on the file structure information of the medical image.

[0505] In other embodiments, the non-image information normality determination step S3715 may include determining whether patient information has been leaked from the medical image acquired from the image acquisition device. Here, patient information may include, but is not limited to, personal information such as the patient's name and age, or information about the patient's health condition.

[0506] For example, if a medical image does not contain patient information, it is unclear which patient the image analysis result from the image analysis device 3000 relates to. Therefore, the non-image information normality determination step S3715 may include determining whether or not patient information is included in the medical image. This may also include obtaining information regarding whether or not patient information is entered into the medical image.

[0507] Figure 22 is a diagram illustrating the first image quality determination unit 3710.

[0508] Referring to Figure 22, the first image quality determination unit 3710 may include at least one of the non-image information acquisition unit 3711 or the non-image information normality determination unit 3717.

[0509] The non-image information acquisition unit 3711 may include a file structure information acquisition unit 3713 or a patient information acquisition unit 3715. The non-image information acquisition unit 3711 can acquire non-image information based on the acquired medical image. In this case, the acquired medical image may include raw data before preprocessing or an image after preprocessing has been performed.

[0510] The non-image information normality determination unit 3717 can determine whether the non-image information is normal based on the non-image information acquired by the non-image information acquisition unit 3711. The non-image information normality determination unit 3717 can output information regarding the result of determining whether the non-image information is normal. For example, the non-image information normality determination unit 3717 can output information regarding whether image analysis is performed normally based on the file structure information of the medical image, or information regarding whether patient information is entered into the medical image.

[0511] In one embodiment, the second image quality determination unit 3730 can perform the function of extracting noise information contained in the image. Here, noise information can mean various types of defects or conditions that affect the acquisition of information based on the image. For example, noise information may include information related to the resolution of the image, information related to the brightness of the image, or artifact information that has occurred in the image, and may also mean various image defects created as a result of inappropriate video sampling.

[0512] Figure 23 is a diagram illustrating the second image quality judgment process.

[0513] Referring to Figure 23, the second image quality determination process may include a medical image acquisition step S3731, a step S3733 for determining whether or not artifacts occur in the medical image, a step S3735 for acquiring artifact information from the medical image, and an artifact information output step S3737.

[0514] Step S3733, which determines whether or not artifacts occur in the medical image, may include determining whether or not artifacts are present in the medical image. Here, artifacts may include noise associated with pixels that do not faithfully represent anatomical structures in the medical image. Exemplary examples of artifacts may include, but are not limited to, motion artifacts, bias artifacts, zipper artifacts, ghost artifacts, and spike artifacts, and may include various types of artifacts that are already publicly known.

[0515] Step S3735, which involves obtaining artifact information from a medical image, may also include obtaining information about artifacts that have occurred within the medical image. Here, the artifact information may include information related to the artifact, such as information about whether or not an artifact has occurred, information about the location of the artifact, information about the type of artifact that occurred, and information about the extent of the artifact's occurrence.

[0516] In one embodiment, the step S3733 for determining whether or not an artifact has occurred in the medical image, or the step S3735 for acquiring artifact information from the medical image, may be performed using a learned neural network model. The step S3733 for determining whether or not an artifact has occurred in the medical image may be performed using a learned neural network model. In the drawing, the step S3733 for determining whether or not an artifact has occurred in the medical image and the step S3735 for acquiring artifact information from the medical image are shown separately, but each step may be performed using a single neural network model to determine whether or not an artifact has occurred and to acquire information related to the artifact.

[0517] The neural network model can determine whether or not artifacts have occurred based on medical images. The neural network model can determine whether artifacts exist within a specific region of the acquired medical image, based on the input medical image.

[0518] Furthermore, the neural network model can acquire artifact information based on medical images. Based on the input medical image, the neural network model can acquire information regarding the presence or absence of artifacts, the location of artifacts, the type of artifacts that occurred, and the extent of artifact occurrence.

[0519] Artifact information can be obtained based on medical images acquired by an image acquisition device. For example, artifact information can be obtained using a neural network model trained to acquire artifact information based on medical images.

[0520] To obtain artifact information, sorter algorithms can be used to classify or predict input images for one or more labels. Various forms of algorithms can be used for classification or prediction. For example, k-nearest neighbors, support vector machines, artificial neural networks, decision trees, self-organizing maps, and logistic regression can be used.

[0521] Artificial neural networks can include sorters, hybrid classifiers, ensemble classifiers, and linear regression networks. Artificial neural networks can also include Convolutional Neural Networks (CNNs). Artificial neural networks can be supervised, unsupervised, or reinforcement-learned models.

[0522] Figure 25 is a diagram illustrating a neural network model for acquiring artifact information according to one embodiment. Referring to Figure 25, the neural network model according to one embodiment may be provided in sorter form. According to one embodiment, the neural network model includes an input layer (IL), a pooling layer (PL), a convolutional neural network layer (CL), a fully connected layer (FCL), a hidden layer (HL), and an output layer (OL), and can acquire feature vectors based on the input image. The neural network model may be provided in sorter form that classifies the input image into one or more labels. Alternatively, the neural network model may be provided in regression model form. The neural network model may be provided as a regression model that acquires a linear output value for specific artifact information based on the input image.

[0523] A neural network model can take medical images acquired by photographing specific regions of the human body as input and obtain output information. The output information can indicate whether the input image contains a target object. For example, the output layer of a neural network model can include output nodes with probability functions. The output layer can include output nodes with probability functions indicating whether the target image contains a target object. For one or more target objects, the output layer can include output nodes with one or more probability functions indicating whether the input image contains each respective target object.

[0524] A neural network model can be trained to acquire artifact information. A neural network model can be trained to acquire artifact information based on training data that includes one or more medical images labeled with artifact information. A neural network model can be trained to acquire artifact regions based on training data that includes one or more medical images labeled with artifact regions.

[0525] Artifact training data can include multiple medical images. Artifact training data can include medical images taken in various ways, such as CT, MRI, or X-ray images. Artifact training data can include multiple medical images containing artifacts of various types, extents, sizes, shapes, or locations.

[0526] Artifact training data can include medical images that have artifact labels indicating the presence or absence of artifacts. Artifact training data can be obtained by photographing various parts of the human body and can include medical images that have artifact labels indicating the presence or absence of artifacts. Here, artifact labels can be assigned differently depending on the location of the artifact, the extent of the artifact, the form of the artifact, or the type of artifact.

[0527] Artifact training data may include medical images with artifact occurrence regions masked. Artifact training data may include medical images masked (or labeled) for one or more artifact occurrence regions. Artifact training data may include medical images masked (or labeled) differently for multiple types of artifact occurrence regions. For example, artifact training data may include medical images displayed in a first color for the occurrence of a first type of artifact and in a second color for the occurrence of a second type of artifact.

[0528] The neural network model can be trained using the artifact training data described above. The neural network model can be trained using the artifact training data in a supervised, unsupervised, or reinforcement learning manner to acquire artifact information based on medical images. The neural network model can also be trained using backpropagation.

[0529] A neural network model can be trained to classify medical images based on whether they contain artifacts, using artifact training data that includes medical images labeled for the presence or absence of artifacts. A neural network model can be trained to classify medical images based on the type of artifacts contained within them. A neural network model can be trained to determine whether a target image contains each type of artifact for multiple artifacts, using artifact training data that includes medical images labeled for the presence or absence of multiple types of artifacts.

[0530] A neural network model can be trained to acquire artifact region information via artifact training data. A neural network model can be trained to detect artifact regions from a target image using artifact training data that includes medical images masked for artifact locations for one or more types of artifacts. A neural network model can be trained to acquire artifact region information for each of multiple artifacts contained in a medical image, for the region and / or type in which each artifact is distributed, via artifact training data that includes medical images labeled for multiple types of artifact regions.

[0531] The neural network model can also be trained using multiple artifact training datasets. The neural network model can be trained using a first artifact training dataset containing medical images with a first type of artifact, and a second artifact training dataset containing medical images with a second type of artifact.

[0532] On the other hand, multiple neural network models may be trained and used. The first neural network model may be trained based on the first artifact training data, and the second neural network model may be trained based on the second artifact training data. The second artifact training data may differ from the first artifact training data, at least in part.

[0533] Figure 24 illustrates how the second image quality assessment process uses an artificial neural network model to determine the location of artifacts within a medical image. Referring to Figure 24, the neural network model can obtain artifact region information regarding the location of artifacts by taking a medical image acquired via an image acquisition device as input.

[0534] Referring to Figure 24, step S3735, which involves obtaining artifact information from a medical image, may include obtaining artifact region information within the medical image.

[0535] Artifact region information can be provided as a display indicating areas in a medical image where artifacts are likely to be located, for example, in the form of a heatmap or a saturation map. Artifact region information may be a feature map obtained from a neural network model that determines the presence or absence of artifacts or from a neural network model that acquires artifact information.

[0536] An artifact region is obtained from a trained neural network model and may be a region on the feature map whose association with the target artifact is greater than or equal to a threshold value, based on the feature map associated with the target artifact present in the target medical image. An artifact region may include a first region and a second region located within the first region. The first region may be a region on the feature map whose association with the target artifact is greater than or equal to a first threshold value. The second region may be a region on the feature map whose association with the target artifact is greater than or equal to a second threshold value, which is greater than the first threshold value.

[0537] Referring to Figures 24(a) to (c), step S3735, which involves obtaining artifact information from a medical image, may include obtaining a first-source image (a), a second-source image (b), and a third-source image (c) that contain the artifact. In this case, the first-source image (a) to the third-source image (c) may each be images taken in different planes. The first-source image (a) to the third-source image (c) may be images taken of the same subject at the same time.

[0538] Referring to Figures 24(d) through (f), step S3735, which involves acquiring artifact information from a medical image, may include acquiring a first artifact image (d), a second artifact image (e), and a third artifact image (f) in which the location of the artifacts is represented. Here, the first artifact image (d) may be an image in which artifact region information acquired based on the first source image (a) is displayed on the first source image (a). The second artifact image (e) may be an image in which artifact region information acquired based on the second source image (b) is displayed on the second source image (b). The third artifact image (f) may be an image in which artifact region information acquired based on the third source image (c) is displayed on the third source image (c).

[0539] Furthermore, referring to Figures 24(d) through (f), the first through third artifact images may be images in which the first region A1 through fourth region A4 are displayed in order of the likelihood of the artifact being located within the medical image. Here, the first region A1 through fourth region A4 may correspond to the region in which the artifact is located within the image. The first region A1 through fourth region A4 may include at least a portion of the region in which the artifact is located within the image. The first region A1 may be a region included in the second region A2. The second region A2 may be a region included in the third region A3. The third region A3 may be a region included in the fourth region A4.

[0540] Although not shown in the drawing, artifact region information may include a bounding box to display the region where the artifact is located within the medical image. Artifact region information may include coordinate information, pixel information, etc., of the location of the artifact within the medical image. For example, the second image quality determination unit 3730 can use artificial neural network models such as BBMs (Backpropagation Based Method), ABMs (Activation Based Method), and PBMs (Pertubation Based Method) to acquire artifact occurrence location information. Here, BBMs methods may include LRP (Layer-wise Relevance Propagation), DeepLIFT, SmoothGrad, VarGrad, etc., ABMs methods may include CAM (Class Activation Map), Grad-CAM (Gradient-Class Activation Map), etc., and PBMs methods may include LIME (Local Interpretable Model-Agnostic Explanation), etc.

[0541] In one embodiment, the third image quality determination unit 3750 can perform a quality determination based on segmentation information obtained by anatomically segmenting a medical image acquired from an image acquisition device, taking into consideration whether the segmentation information satisfies quantitative criteria. As a more specific example, the third image quality determination unit 3750 can determine whether at least some of the anatomical structures contained in the medical image satisfy predetermined quantitative criteria.

[0542] The third image quality determination unit 3750 can determine whether a value related to a predetermined area corresponding to a part of the human body included in the image satisfies a standard. The third image quality determination unit 3750 can determine whether at least a portion of the anatomical structure of the human body captured via the image acquisition device has been completely captured.

[0543] For example, when analyzing images of specific parts of the human body taken via an imaging device, a disease can be diagnosed based on morphological indicators of at least some of the captured areas of the human body. In this case, if each area of ​​the human body is not fully captured, the diagnosis cannot be made with a certain level of reliability. Therefore, since each area of ​​the human body should be fully captured, the third image quality determination unit 3750 can perform a function to determine whether each area of ​​the human body has been fully captured by considering whether the morphological indicators based on the anatomical areas of the human body included in the medical image satisfy generally required standards.

[0544] As a more specific example, when the image analysis device 3000 analyzes MRI images of the brain, it can calculate the ratio of volume values ​​for each anatomical region of the brain to determine a disease. In this case, if the volume value of a specific region of the brain that has been imaged does not meet the generally required standard value, the disease diagnosis cannot have a certain level of reliability. In this case, if a specific region of the brain is not fully imaged, and the disease diagnosis is made based on the volume value of that region, the result of the disease diagnosis may be inaccurate. Therefore, the third image quality determination unit 3750 can perform a function to determine whether each region of the brain has been fully imaged by considering whether the morphological indicators for each region of the brain contained in the MRI image meet the generally required standard value.

[0545] Figure 26 is a diagram illustrating the third image quality judgment process.

[0546] Referring to Figure 26, the third image quality judgment process may include an image segmentation information acquisition step S3751, a quality judgment target area acquisition step S3753, a morphological index acquisition step S3755 for the target area, and a step S3757 for comparing the morphological index of the target area with a reference value.

[0547] The image segmentation information acquisition step S3751 may include acquiring segmentation information acquired based on medical images segmented by the image segmentation module 3500.

[0548] Segmentation information may include information about the anatomical structure of the human body obtained by segmenting medical images. Segmentation information may include morphological indicators of at least some regions of the anatomical structure of the human body obtained by segmenting medical images. Furthermore, segmentation information may include morphological values ​​of at least some regions of the anatomical structure of the human body obtained by segmenting medical images.

[0549] Morphological indices can be obtained based on morphological values. Morphological indices can be obtained based on multiple morphological values, for example, a first morphological value and a second morphological value. For example, a morphological indice can be obtained based on the ratio of the first morphological value to the second morphological value. In this case, the first morphological value may be the cerebellum, the second morphological value may be an anatomical region within the skull, and the morphological indice obtained by the ratio of the first and second morphological values ​​may represent the ICV value. Morphological values ​​may be information regarding the area, volume, location, or shape of at least some regions of the anatomical structures of the human body.

[0550] Step S3753, which involves acquiring the target area for quality judgment, may include identifying the target area for performing a third image quality judgment based on segmentation information. Here, the target area may include at least a portion of the anatomical structures of the human body acquired through medical image segmentation. Furthermore, the target area may include at least a portion of the anatomical structures of the human body that can influence the reliability of the disease judgment result.

[0551] For example, the target region may include the outermost region of the anatomical structure of the human body obtained through medical image segmentation, located at the center of the medical image. The target region may include the region of the anatomical structure of the human body obtained through medical image segmentation that forms the basis for disease diagnosis. The target region may include any one of the leftmost, rightmost, uppermost, or lowermost regions of the anatomical structure of the human body obtained through medical image segmentation. When image quality assessment is performed based on medical images of the brain, the target region may include at least a portion of the internal cranial region of the brain obtained through medical image segmentation.

[0552] Step S3755 for obtaining morphological indicators of a target area may include steps to obtain morphological indicators relating to the identified target area. For example, step S3755 for obtaining morphological indicators of a target area may include obtaining morphological indicators of the target area based on at least some of a plurality of medical images secured by an image acquisition device. Step S3755 for obtaining morphological indicators of a target area may include obtaining morphological values ​​of the target area based on at least some of a plurality of medical images secured by an image acquisition device.

[0553] Step S3757, which compares morphological indicators of the target area with reference values, may include comparing morphological indicators or morphological values ​​of the target area with predetermined reference values. Here, predetermined reference values ​​mean the average value of morphological values ​​of the target area for a large number of general individuals. Predetermined reference values ​​may mean indicators or measured values ​​related to the target area when it is possible to obtain medical information with a certain level of reliability or higher based on the morphological indicators of the target area when diagnosing a disease.

[0554] Figure 27 is a diagram illustrating one embodiment of the third image quality judgment process.

[0555] Referring to Figures 27(a) to (c), the third image quality judgment process according to one embodiment may include performing a quality judgment based on medical images relating to the brain. According to one embodiment, the third image quality judgment process may include performing a quality judgment based on segmentation information obtained by anatomically segmenting medical images relating to the brain, regarding whether the segmentation information meets quantitative criteria.

[0556] The image segmentation information acquisition step S3751 may include receiving segmentation information obtained by anatomically segmenting an MRI image of the brain. The quality judgment target area acquisition step S3753 may include identifying a quality judgment target area from among the anatomically segmented brain regions, for example, a predetermined region corresponding to the cerebellum from among the anatomically segmented brain regions.

[0557] For example, if the region subject to quality assessment is a predetermined region corresponding to the cerebellum, the segmentation information received in the image segmentation information acquisition step S3751 may be information acquired based on an MRI image in which a predetermined region corresponding to one part of the brain was not completely captured. Here, the criterion for whether the cerebellum was completely captured can be applied as the same criterion as the criterion for defining the internal cranial region based on the morphological values ​​of the cerebellum in the part describing ICV of this specification.

[0558] Referring to Figure 27(a), the image segmentation information acquisition step S3751 may include receiving segmentation information acquired based on a first target region image in which at least a portion of a predetermined region corresponding to the cerebellum among the various regions of the brain was not captured.

[0559] Referring to Figure 27(b), the image segmentation information acquisition step S3751 may include receiving segmentation information acquired based on a second target region image in which all predetermined regions corresponding to the cerebellum among the various regions of the brain have been captured.

[0560] Referring to Figure 27(c), the image segmentation information acquisition step S3751 may include receiving segmentation information acquired based on a third target region image in which a predetermined region corresponding to the cerebellum among the various regions of the brain has been fully imaged, but at least a portion of the regions corresponding to other anatomical structures other than the cerebellum has not been fully imaged.

[0561] Step S3755, which involves obtaining morphological indicators for the target region, may include obtaining morphological indicators for the region subject to quality judgment, such as the volume value of the cerebellum identified in the region subject to quality judgment. Step S3755, which involves obtaining morphological indicators for the target region, may also include comparing the morphological indicators for the target region with a reference value. Step S3755, which involves obtaining morphological indicators for the target region, may also include comparing the volume value of the cerebellum identified in the target region with a predetermined reference value to determine whether or not a quantitative criterion is met.

[0562] For example, a predetermined reference value would mean a morphological index of the cerebellum obtained based on a second target region image in which a predetermined region corresponding to the cerebellum is completely captured, as shown in Figure 27(b). A predetermined reference value can be obtained based on a morphological index of the cerebellum obtained based on multiple images in which a predetermined region corresponding to the cerebellum is completely captured. A predetermined reference value may be the average value of the cerebellar volume obtained based on multiple images.

[0563] In this case, step S3757, which compares the morphological indicators of the target region with a reference value, may include comparing the morphological indicators of the cerebellum obtained based on a first target region image in which at least a portion of a predetermined region corresponding to the cerebellum was not captured, as shown in Figure 27(a), with the morphological indicators of the cerebellum obtained based on a second target region image in Figure 27(b), i.e., a predetermined reference value, to determine whether a quantitative criterion is satisfied. Step S3757, which compares the morphological indicators of the target region with a reference value, may also include performing an image quality judgment based on whether the volume value of the cerebellum obtained based on the first target region image in Figure 27(a) shows a difference of a certain range or more from the predetermined reference value.

[0564] Referring to Figure 27(c), in the case of a third target region image where the region corresponding to the cerebellum identified as the target region is fully imaged, but at least part of the regions corresponding to other anatomical structures other than the cerebellum is not fully imaged, the third image quality assessment process may include re-identifying the target region for quality assessment and then performing the image quality assessment.

[0565] Referring to Figure 27(c), when image quality assessment is performed based on the region corresponding to the cerebellum, it is possible to assess leakage in the inferior part of the image, but it can be difficult to assess leakage in the superior part of the image. To avoid this problem, other indicators besides the region corresponding to the cerebellum can be used as indicators for image quality assessment. For example, the volume (or ICV) of the region corresponding to the skull or cerebrum can be used as an indicator for quality assessment.

[0566] Alternatively, if morphological indicators obtained based on the region corresponding to the cerebellum (e.g., the ICV value of the cerebellum) are above the reference value, it can be determined that imaging of regions other than the cerebellum was incomplete.

[0567] In one embodiment, the fourth image quality determination unit 3770 can perform image quality determination based on the relationship between segmentation information obtained by anatomically segmenting a medical image acquired from an image acquisition device and artifact information extracted from the medical image acquired from the image acquisition device. For example, the fourth image quality determination unit 3770 can determine whether artifacts occurring in a medical image overlap with at least a portion of the anatomical structures included in the medical image.

[0568] For example, when the image analysis device 3000 analyzes a specific part of the human body after capturing it via an image acquisition device, it can determine a disease based on segmentation information obtained by anatomically segmenting each region of the human body. In this case, if an artifact occurs in the segmentation information that can serve as the basis for the disease determination, the result of the disease determination based on the medical image cannot have a reliability level above a certain level. Conversely, even if an artifact occurs in the medical image, if the artifact does not affect the segmentation information that can serve as the basis for the disease determination, for example, if the artifact occurs in a region that does not overlap with the segmentation information, the result of the disease determination based on the medical image can have a reliability level that satisfies a certain level. Therefore, even if an artifact occurs in the medical image, the fourth image quality determination unit 3770 can perform a function to determine whether the medical image can be analyzed normally by considering whether the artifact affects the segmentation information that can serve as the basis for the disease determination.

[0569] As a more specific example, when the image analysis device 3000 analyzes an MRI image of the brain, it can determine a disease based on segmentation information obtained by anatomically segmenting each region of the brain. In this case, if artifacts occur in at least a portion of the segmented anatomical regions of the brain that can serve as the basis for the disease determination, the result of the disease determination based on the medical image cannot have a certain level of reliability. Therefore, the fourth image quality determination unit 3770 can perform a function to determine whether or not the medical image can be analyzed normally by considering whether or not artifacts have occurred in at least a portion of the anatomical structures of each region of the brain included in the MRI image.

[0570] Figure 28 is a diagram illustrating the fourth image quality judgment process.

[0571] Referring to Figure 28, the fourth image quality determination process may include an artifact location information acquisition step S3771, an image segmentation information acquisition step S3773, a region of interest acquisition step S3775, and a relationship determination step S3777 between the artifact location and the region of interest.

[0572] The artifact occurrence location information acquisition step S3771 may include acquiring artifact occurrence location information. At this time, the artifact occurrence location information may be acquired from the image analysis device 3000. The artifact occurrence location information may include artifact region information relating to the artifact occurrence location, for example, a saturation map in the form of a heat map showing a region in the medical image where an artifact is likely to be located, a bounding box for displaying the region in the medical image where an artifact is located, coordinate information of the location of the artifact in the medical image, or pixel information. A detailed explanation of this has been given above in relation to the second image quality determination unit 3730, so a detailed explanation is omitted here.

[0573] The artifacts described below may include noise, for example, various types of defects or conditions that affect the acquisition of information based on an image. Artifacts may also include image resolution, image brightness, and various other image defects resulting from improper video sampling.

[0574] The image segmentation information acquisition step S3773 may include acquiring segmentation information acquired based on a medical image segmented by the image segmentation module 3500. Here, the image segmentation information may include information acquired by anatomically segmenting a medical image. The image segmentation information may include segmentation information acquired by anatomically segmenting an MRI image of the brain, and a detailed explanation of this has been given above in relation to the third image quality determination unit 3750, so a detailed explanation is omitted here.

[0575] Step S3775 for acquiring the region of interest may include acquiring information related to the region of interest based on segmentation information obtained by anatomically segmenting a medical image. Step S3775 for acquiring the region of interest may also include acquiring information related to the region of interest from among the segmentation information obtained by anatomically segmenting a medical image. Step S3775 for acquiring the region of interest may acquire one or more regions by segmenting a medical image and acquiring the region corresponding to the region of interest.

[0576] The region of interest may include at least some regions of the anatomical structures of the human body acquired through medical image segmentation. The region of interest can be predefined as a region corresponding to a specific anatomical (or functional) element in the medical image. The region of interest may include the regions of the anatomical structures of the human body acquired through medical image segmentation that are located furthest from the center of the medical image. The region of interest may include regions of the anatomical structures of the human body acquired through medical image segmentation that form the basis for disease diagnosis. When image quality assessment is performed based on MRI images of the brain, the region of interest may include at least some of the cranial and intracranial regions of the brain acquired through medical image segmentation. The region of interest may be an element related to the diagnosis of the target disease.

[0577] Step S3777, which determines the relationship between the artifact location and the region of interest, may include determining the relationship between the artifact location and the region of interest. Step S3777 may also include performing a quality judgment based on information about artifacts contained in the medical image and information about the region of interest among the anatomical structures of the human body obtained by segmenting the medical image. Step S3777 can perform an image quality judgment based on whether the degree of overlap between artifacts occurring in the medical image and the region of interest affects the reliability of the disease diagnosis below a certain level.

[0578] Step S3777, which determines the relationship between the artifact occurrence location and the region of interest, may include performing a quality determination based on whether at least a portion of the artifacts that occurred in the medical image are located on the region of interest, or whether at least a portion of the artifacts that occurred in the medical image overlap with the region of interest.

[0579] Step S3777, which determines the relationship between the artifact location and the region of interest, may include determining the relationship between the artifact location and the region of interest based on the relationship between the artifact boundary and the region of interest. Here, the artifact boundary may be the outline of the artifact region. The artifact boundary may be obtained from a neural network model that determines the presence or absence of artifacts, and may be obtained based on a spleness map (e.g., CAM) associated with the artifact location estimation.

[0580] Step S3777, which determines the relationship between the artifact location and the region of interest, may include obtaining the outlines of the artifact region and the region of interest on the medical image, and determining whether the artifact and the region of interest overlap based on whether the outlines of the artifact region and the region of interest overlap on the medical image. In this case, the image analyzer 3000 can determine that the quality of the medical image is normal if there are fewer than two intersections where the outlines of the artifact region and the region of interest overlap on the medical image. Alternatively, the image analyzer 3000 can determine that the quality of the target medical image is abnormal if there are two or more intersections.

[0581] Step S3777, which determines the relationship between the artifact location and the region of interest, may include obtaining the number of common pixels included in both the artifact region and the region of interest on the medical image, and using the number of common pixels to determine the degree of overlap between the artifact and the region of interest. In this case, the image analyzer 3000 can determine that the quality of the medical image is abnormal if the number of common pixels exceeds a predetermined standard value. Alternatively, the image analyzer 3000 can determine that the quality of the medical image is normal if the number of common pixels is less than or equal to a predetermined standard value.

[0582] Step S3777, which determines the relationship between the artifact occurrence location and the region of interest, may include determining whether the artifact occurred in at least a portion of the region of interest, for example, the skull region or the internal skull region. Step S3777, which determines the relationship between the artifact occurrence location and the region of interest, may also include determining whether the artifact overlaps with at least a portion of the region of interest, for example, the skull region and the internal skull region.

[0583] Figure 29 is an example of an image in which artifacts have occurred. Below, the image quality assessment process, along with the extent of artifact occurrence, will be explained with reference to Figures 28 and 29.

[0584] Referring to Figure 29(a), an image can be obtained that includes artifacts overlapping the skull region and the internal skull region. Referring to Figure 29(b), an image can be obtained that includes artifacts overlapping a portion of the skull region. Referring to Figures 29(a) and (b), the region of interest may be the internal skull region or the region corresponding to the skull.

[0585] Referring to Figure 29(a), if the target medical image includes artifacts that overlap with the cranial region and the internal region of the skull, and the region of interest is the internal region of the skull or the cranial region, determining the relationship between the artifact occurrence location and the region of interest (S3777) may include determining that the region of interest and the artifact occurrence location overlap.

[0586] Referring to Figure 29(b), if the target medical image includes artifacts that overlap with the skull region, and the region of interest is the internal skull region, determining the relationship between the artifact location and the region of interest (S3777) may include determining that the region of interest and the artifact location do not overlap.

[0587] Referring to Figure 29(b), if the target medical image includes artifacts that overlap with the skull region, and the region of interest is the skull region, determining the relationship between the artifact occurrence location and the region of interest (S3777) may include determining that the region of interest and the artifact occurrence location overlap.

[0588] The artifacts illustrated in Figure 29 represent regions identified as artifact regions. These regions can be obtained through a segmentation neural network model that acquires artifact regions or a sorter model that acquires artifact information. The details of this process will be explained in more detail below.

[0589] The fourth image quality determination may include obtaining multiple reference regions associated with artifacts and determining image quality based on the overlap between each reference region and the region of interest.

[0590] Step S3777, which determines the relationship between the artifact occurrence location and the region of interest according to one embodiment, may include performing an image quality judgment based on artifact information related to the location of the artifact and the region in which the region of interest is distributed. Step S3777, which determines the relationship between the artifact occurrence location and the region of interest, may also include performing an image quality judgment based on information regarding the probability that an artifact is located in a specific region within the medical image and information regarding the region of interest. Step S3777, which determines the relationship between the artifact occurrence location and the region of interest, may also include determining whether the degree of overlap between the artifact occurring in the medical image and the region of interest exceeds a predetermined threshold value.

[0591] Information regarding the probability of an artifact being located in a specific region within a medical image may include artifact region information. Here, artifact region information may include a first artifact region to the nth artifact region, in order of the likelihood of artifacts being located within the medical image. The artifact region may include a first region and a second region located within the first region, as described above in the second image quality judgment process. The first region may be a region on the feature map whose association with the target artifact is greater than or equal to a first criterion value, and the second region may be a region on the feature map whose association with the target artifact is greater than or equal to a second criterion value which is greater than the first criterion value.

[0592] Figure 30 illustrates how, in a fourth image quality assessment process according to another embodiment, it is determined whether an artifact occurred overlapping with the region of interest.

[0593] Referring to Figure 30(b), the artifact region may include the first artifact region AR1, the second artifact region AR2, the third artifact region AR3, or the fourth artifact region AR4, in order of likelihood of the artifact being located within the medical image. The first to fourth artifact regions AR1 through AR4 may include at least a portion of the region in which the artifact is located within the image. The artifact regions may be the first artifact region AR1, the second artifact region AR2, the third artifact region AR3, and the fourth artifact region AR4, in order of their high relevance to the artifact.

[0594] Alternatively, an artifact region may include regions on the feature map whose relevance to the target artifact is greater than or equal to a first criterion (e.g., the third artifact region AR3 and the fourth artifact region AR4). An artifact region may also include regions on the feature map whose relevance to the target artifact is greater than or equal to a second criterion (e.g., the first artifact region AR1 and the second artifact region AR2).

[0595] Referring to Figure 30(b), the degree of overlap between the artifact and the region of interest can be determined by using the number of common pixels included in both the artifact region and the region of interest. For example, if the number of common pixels included in both the third artifact region A3 and the fourth artifact region A4 and the region of interest exceeds a predetermined threshold, the image analyzer 3000 can determine that the quality of the medical image is abnormal. As another example, if the number of common pixels included in both the first artifact region A1 and the second artifact region A2 and the region of interest is less than or equal to a predetermined threshold, the image analyzer 3000 can determine that the quality of the medical image is normal.

[0596] Referring to Figure 30(c), the outline of the artifact region can be used as the artifact boundary. The artifact boundary can be determined based on a feature map associated with the artifact region, as shown in Figure 30(a). The artifact boundary can be determined by the boundary of any one of the first artifact region AR1 to the fourth artifact region AR4, as shown in Figure 30(b). The artifact boundary can be determined by the average of the boundaries of the first artifact region AR1 to the fourth artifact region AR4.

[0597] The artifact boundary may be the outline of the first region (for example, the region on the feature map where the relevance to the target artifact is greater than or equal to the first criterion). The artifact boundary may also be the outline of the second region (for example, the region on the feature map where the relevance to the target artifact is greater than or equal to the second criterion, which is greater than the first criterion).

[0598] Determining the relationship between artifact occurrence locations and regions of interest based on the relationship between artifact boundaries and regions of interest may include performing quality assessments based on whether or not there is overlap between artifact boundaries and regions of interest.

[0599] As shown in Figure 30(c), if there are two or more intersections where the artifact boundary and the outline of the region of interest overlap on the medical image, the image analysis device 3000 can determine that the quality of the target medical image is abnormal.

[0600] If multiple artifacts exist within the target image, the step S3777 for determining the relationship between the artifact location and the region of interest may include performing a quality determination based on whether or not there is overlap or the degree of overlap between one of the multiple artifacts and the region of interest.

[0601] The image output device may include an image quality-related information output module 3900. The image quality-related information output module 3900 can output the image quality judgment results performed by the image quality judgment module 3700 of the image analysis device 3000. In addition, the image quality-related information output module 3900 can output information generated based on the image quality judgment results performed by the image quality judgment module 3700 of the image analysis device 3000.

[0602] An image output device or image analysis device can obtain user input to instruct additional operations in response to the output of a selection window.

[0603] Figure 31 is a diagram illustrating an image quality-related information output module according to one embodiment.

[0604] Referring to Figure 31, the image quality-related information output module 3900 may include at least one of either an error information output unit 3910 or a selection window output unit 3930. In this case, the error information output unit 3910 can output error information related to the information acquired based on the image quality judgment result. The selection window output unit 3930 can output a selection window related to the information acquired based on the image quality judgment result.

[0605] The error information output unit 3910 and the selection window output unit 3930 can each output information determined based on independent image quality judgment results. For example, the error information output unit 3910 can output error information determined based on a first quality judgment result, and the selection window output unit 3930 can output a selection window determined based on a second quality judgment result.

[0606] Alternatively, the error information output unit 3910 and the selection window output unit 3930 can output information determined based on the same image quality judgment result. The error information output unit 3910 can output error information determined based on the first quality judgment result, and the selection window output unit 3930 can output a selection window determined based on the first quality judgment result.

[0607] In Figure 31, the error information output unit 3910 and the selection window output unit 3930 are shown as separate components; however, this is merely an example, and the error information output unit 3910 and the selection window output unit 3930 may be provided in a single physical or logical configuration. For example, the image quality-related information output module 3900 can also output error information generated based on the quality judgment result and a selection window corresponding to the error information.

[0608] The selection window output unit 3930 can output a selection window for obtaining user instructions related to additionally executable actions based on the quality judgment result.

[0609] The image quality-related information output module 3900 can output at least one of the following: error information or a selection window, in text or image format.

[0610] The image quality-related information output module 3900 can output error information. In this case, the error information may include information obtained based on the image quality judgment result performed by the image quality judgment module 3700. The error information may also include at least one of the following: first error information associated with the first image quality judgment result, second error information associated with the second image quality judgment result, third error information associated with the third image quality judgment result, and fourth error information associated with the fourth image quality judgment result.

[0611] The image quality-related information output module 3900 can output a first error information screen. The first error information may include information obtained based on the first image quality judgment result. The first error information may also include information obtained based on the metadata information of the medical image.

[0612] According to one embodiment, the image quality-related information output module 3900 can output a first error information screen if there is an abnormality in the file structure information of the medical image acquired from the image acquisition device, such as the file format of the medical image, the format of the medical image, or the file size. In this case, the first error information screen may include a first notification window indicating that there is an abnormality in the file structure information of the medical image, or information regarding the file structure of the medical image.

[0613] According to another embodiment, the image quality-related information output module 3900 can output a first error information screen if patient information, such as personal information related to the patient's name and age, information related to the time the medical image was taken, or information related to the patient's health status, is missing from the medical image acquired from the image acquisition device. In this case, the first error information may include a first notification window indicating that patient information is missing from the medical image, or patient information that is missing from or included in the medical image.

[0614] Figure 32 is a diagram illustrating the first error information screen.

[0615] Referring to Figure 32(a), the first error information screen may include at least one of the following: medical image MI based on the first image quality judgment, medical image file structure information DI, or patient information PI entered into the medical image.

[0616] In this case, the medical image MI may be the image displaying the medical image used for the first image quality determination. The medical image file structure information DI may include first file structure information regarding the format of the medical image, second file structure information regarding the format of the medical image, or third file structure information regarding the size of the medical image. In this case, although not shown in the drawing, the first error information screen includes a detailed information object, and the output screen may additionally display more detailed information about the medical image MI or file structure information DI (e.g., patient's date of birth, patient's address, time of shooting, location of shooting, shooting device, etc.) based on user selection for the detailed information object.

[0617] Referring to Figure 32(b), the first error information screen may include a first notification window CO containing the first information that the first image quality judgment result indicates an abnormality in the medical image. Illustratively, the first notification window CO may include an error code indicating an abnormality in the medical image's file structure or that patient information has been lost from the medical image. In this case, the first error information screen may include a detailed information object, and the output screen may additionally display detailed information about the first information (e.g., an explanation of the error code, the cause of the error message, a manual that can address the error message) based on user selection of the detailed information object.

[0618] The image quality-related information output module 3900 can output a second error information screen. The second error information may include information obtained based on the second image quality judgment result. The second error information may include information obtained based on noise information. The second error information may include information obtained based on various types of defects or conditions that affect the acquisition of information based on the image.

[0619] According to one embodiment, the image quality-related information output module 3900 can output a second error information screen if the medical image acquired from the image acquisition device contains noise. The image quality-related information output module 3900 can output the second error information in text or image format.

[0620] For example, the image quality-related information output module 3900 can output a second error information screen if the medical image acquired from the image acquisition device contains artifacts. The second error information screen may include an error message indicating that the medical image contains artifacts. The second error information may also include one of the following: information about the location where the artifact occurred in the medical image, information about the extent of the artifact, information about the range of the artifact, or information about the type of artifact.

[0621] The second error information screen may contain text or images. For example, the second error information screen may include a display to indicate areas in the medical image where artifacts are likely to be located, such as a heatmap image. Alternatively, the second error information screen may include an image representing a bounding box indicating the location of artifacts in the medical image. Another example is that the second error information screen may include an image to show anatomical segmentation information in the medical image.

[0622] Figures 33 and 34 are illustrative diagrams illustrating the second error information screen.

[0623] Referring to Figure 33(a), the second error information screen may include a medical image MI display area related to the second image quality judgment. The second error information screen may also include an artifact information AF output area.

[0624] Medical image MI may be an image in which areas highly likely to contain artifacts are highlighted based on a second-level image quality assessment. Medical image MI may be an image in which areas judged to be highly associated with artifacts are highlighted in a heatmap format. Medical image MI can be zoomed in or out by the user to allow for a more detailed examination of the type, extent, and form of artifacts.

[0625] Artifact information AF may include at least one of the following: information about the type of artifact, information about the region where the artifact is likely to be located, or information about the extent of the artifact's occurrence. The second error information screen includes a detailed information object, and the output screen may additionally display more detailed information about the medical image MI or artifact information AF (e.g., information about the probability that an artifact is located in a specific region of the image, heatmap information about the extent of the artifact's occurrence, or boundary information) depending on the user's selection of the detailed information object.

[0626] Referring to Figure 33(b), the second error information screen may include a second notification window CO containing second information about artifacts contained within the medical image. For example, the second notification window CO may include a message displaying an error code indicating that an artifact has occurred within the medical image. In this case, the second error information screen may include a detailed information object, and the output screen may additionally display more detailed information about the second information (e.g., an explanation of the error code, the cause of the error message, a manual that can address the error message, etc.) based on the user's selection of the detailed information object.

[0627] Referring to Figure 34, artifact information AF may include first artifact information AF1 containing information about the type of artifact, second artifact information AF2 containing information about the region where the artifact is likely to be located, or third artifact information AF3 containing information about the extent to which the artifact occurs.

[0628] The first artifact information AF1 may include information about the type of artifact, information about whether the image in which the artifact occurred can be corrected, and so on. In this case, the information about whether the image in which the artifact occurred can be corrected may include information obtained based on any one of the following: the type of artifact, the location of occurrence, the extent of occurrence, or the degree of occurrence.

[0629] The second artifact information AF2 may include information about areas in the medical image where artifacts are likely to be located, and information about analyzable areas among the areas corresponding to anatomical structures in the medical image. In this case, information about analyzable areas can be obtained based on anatomical segmentation information and artifact occurrence location information. For example, information about the diagnostic area may include information about areas in the medical image corresponding to anatomical structures that do not overlap with artifacts, or information about areas in the medical image corresponding to anatomical structures that have a low degree of overlap with artifacts. In this case, the diagnosable diseases may differ depending on the analyzable area.

[0630] The third artifact information AF3 may include information regarding the extent, level, and range of the artifact. In this case, the image analysis device can make a decision on whether or not to proceed with image analysis based on the third artifact information AF3. For example, even if an artifact occurs in a medical image, the image analysis device can decide to proceed with image analysis if it determines, based on the third artifact information AF3, that the extent of the artifact is lower than a predetermined threshold value.

[0631] The image quality-related information output module 3900 can output a third error information screen. The third error information may include information obtained based on the third image quality judgment result. In addition, the third error information may include information obtained based on segmentation information obtained by anatomically segmenting medical images acquired from the image acquisition device.

[0632] According to one embodiment, the image quality-related information output module 3900 can output a third error information screen if at least some of the anatomical structures included in the medical image acquired from the image acquisition device do not satisfy predetermined quantitative criteria. For example, the image quality-related information output module 3900 can output a third error information screen if a value related to a predetermined region corresponding to a part of the human body included in the medical image does not satisfy the criteria. As a more specific example, the image quality-related information output module 3900 can output a third error information screen if the volume value of a predetermined region corresponding to a part of the brain included in an MRI image of the brain does not satisfy a predetermined criterion value. Furthermore, the image quality-related information output module 3900 can output the third error information screen in text or image format.

[0633] Third-party error information may include an error message indicating an abnormality in at least some of the anatomical structures within the medical image, or information about the abnormal anatomical structures. More specifically, third-party error information may include information related to the target area for which quality judgments are made based on the image segmentation results. Third-party error information may include information related to at least some of the anatomical structures of the human body obtained through image segmentation that can affect the reliability of the disease judgment results.

[0634] The third error information may include morphological indicators or morphological values ​​of at least some regions of the anatomical structures of the human body obtained by segmenting the medical image, such as information on area, volume, location, or shape.

[0635] Third-party error information may include information regarding the results of comparing morphological indicators or morphological values ​​of the target area with predetermined reference values. More specifically, third-party error information may include information regarding whether at least some of the anatomical structures included in the medical image meet predetermined quantitative criteria.

[0636] The third error information screen may include text or images. The third error information screen may include text that numerically represents morphological indicators or morphological values ​​of at least some regions of the anatomical structures of the human body obtained by segmenting the medical image. The third error information screen may include an image in which the target region corresponding to the anatomical structure that serves as the criterion for quality judgment is displayed on the original medical image.

[0637] Figures 35 and 36 are illustrative diagrams illustrating the third error information screen.

[0638] Referring to Figure 35(a), the third error information screen may include the third image quality judgment and related medical image MI or region information RI.

[0639] In this case, the medical image MI may include information related to the target area used for the third image quality assessment based on the image segmentation results. For example, when determining whether the lower part of a medical image has been completely captured, the medical image may include an image representing information related to the target area on which the quality assessment is performed, such as the cerebellum. The third error information screen may include buttons to zoom in or out on the medical image MI to examine in more detail information about the area corresponding to the anatomical structure within the medical image. The area information RI may also include one of the following: information related to the target area on which the image quality assessment is performed, information about parts that were not completely captured in the medical image, or information about whether the target area meets predetermined quantitative criteria. At this time, the third error information screen may include a detailed information object, and the output screen may additionally display more detailed information about the medical image MI or area information RI (e.g., information about the human body area that was the subject of the quality assessment, information about the image area that was the subject of the quality assessment, etc.) depending on the user's selection of the detailed information object.

[0640] Referring to Figure 35(b), the third error information screen may include a third notification window CO containing third information about the region corresponding to an anatomical structure in the medical image. Exemplary, the third notification window CO may include error information about the region corresponding to an anatomical structure in the medical image, such as a message displaying an error code indicating that the segmentation information did not meet predetermined quantitative criteria. In this case, the third error information screen may include a detailed information object, and the output screen may, at the user's selection of the detailed information object, additionally display detailed information about the third information (e.g., an explanation of the error code, the cause of the error message, a manual that can address the error message, etc.).

[0641] Referring to Figure 36, the region information RI may include first region information RI1, which includes a first target region included in the first part of the medical image and information on whether the first target region meets predetermined quantitative criteria; second region information RI2, which includes a second target region included in the second part of the medical image and information on whether the second target region meets predetermined quantitative criteria; or third region information RI3, which includes information on the arrangement of anatomical structures within the medical image and information on whether correction is possible.

[0642] For example, the first domain information RI1 may include information regarding whether morphological indicators or morphological values ​​relating to a target region located above the medical image, such as the frontal lobe, meet predetermined quantitative criteria.

[0643] The second domain information, RI2, may include information regarding whether morphological indicators or morphological values ​​related to the target region located beneath the medical image, such as the cerebellum, meet predetermined quantitative criteria.

[0644] The third domain information RI3 may include information regarding the arrangement of anatomical structures within the medical image, such as information regarding whether the anatomical structures in the medical image were captured at an angle, and judgment information regarding whether the medical image can be corrected based on this information.

[0645] The image quality-related information output module 3900 can output a fourth error information screen. The fourth error information may include information obtained based on the fourth image quality judgment result.

[0646] According to one embodiment, the image quality-related information output module 3900 can output a fourth error information screen if an artifact occurring in the medical image overlaps with at least a portion of the anatomical structures included in the medical image.

[0647] The fourth error information may include information obtained based on the relationship between segmentation information, which is obtained by anatomically segmenting medical images acquired from an image acquisition device, and artifact information extracted from medical images acquired from an image acquisition device. For example, the fourth error information may include information on whether artifacts occurred in at least some of the anatomical structures of each region of the brain contained in the MRI image.

[0648] The fourth error information may include information on whether artifacts occurring within the medical image overlap with at least some of the anatomical structures contained within the medical image. Furthermore, the fourth error information may include information on the degree of overlap between artifacts occurring within the medical image and at least some of the anatomical structures contained within the medical image. Additionally, the fourth error information may include information on the reliability of disease diagnosis results based on the analysis of the medical image, based on the degree of overlap between artifacts occurring within the medical image and at least some of the anatomical structures contained within the medical image.

[0649] The fourth error information may include information about the relationship between the probability of an artifact being located in a specific region within the medical image and the region of interest. Furthermore, the fourth error information may include at least one of the following: a first artifact region (a region where artifacts are likely to be located within the medical image), a second artifact region (a region where artifacts are typically located within the medical image), or a third artifact region (a region where artifacts are unlikely to be located within the medical image). The first to third artifact regions can be visually highlighted on the original medical image. The first to third artifact regions can also be displayed overlaid on the original medical image in the form of a heatmap.

[0650] The fourth error information screen may include text or images. For example, the fourth error information screen may include images displaying artifact area information on anatomical structures of the human body obtained by segmenting the medical image, such as a heatmap-type saturation map showing areas where artifacts are likely to be located within the medical image, a bounding box for displaying the area where artifacts are located within the medical image, coordinate information of the artifact's location within the medical image, or pixel information.

[0651] Figures 37 and 38 are illustrative diagrams illustrating the fourth error information screen.

[0652] Referring to Figure 37(a), the fourth error information screen may include at least one of the fourth image quality judgments and related medical image MI or information QI regarding the relationship between artifacts and areas of interest.

[0653] Medical image MI can be an image in which a region of interest is displayed within the image. Medical image MI can be an image in which a region is likely to be located within a specific area of ​​the image.

[0654] Medical image MI can display the degree of overlap between the region of interest and the artifact, based on the artifact's location information. For example, medical image MI can display information related to the overlapping area between the region of interest and the artifact, such as the area, shape, and number of pixels of the overlapping area. Medical image MI can also display the ratio of the region of interest to the area of ​​interest that overlaps with the artifact.

[0655] A medical image MI can be an image that displays the probability of artifacts being located in a specific region within a medical image and the relationships between those regions of interest.

[0656] For example, a medical image MI may be an image in which the areas corresponding to artifacts are visually highlighted on the target medical image. A medical image MI may be an image in which the areas corresponding to artifacts are displayed in a heatmap format on the target medical image, or the boundaries of the areas corresponding to artifacts, such as outlines, are displayed.

[0657] Alternatively, a medical image MI may be an image in which the area where the artifact and the region of interest overlap is visually highlighted. A medical image MI may be an image in which the region corresponding to the artifact is displayed on the target medical image. A medical image MI may be an image in which the boundary line, for example, the outline, of the region corresponding to the artifact is displayed on the target medical image. A medical image MI may be an image in which the region corresponding to the artifact is displayed on the target medical image in a heatmap format.

[0658] Information Quality Indicators (QIs) relating artifacts to areas of interest may include, but are not limited to, information about the type of area of ​​interest, information about overlap between areas of interest and artifacts, or information about quality judgment results based on overlapping information.

[0659] In this case, the fourth error information screen includes a detailed information object, and the output screen can additionally display more detailed information QI regarding the relationship between medical image MI or artifacts and the region of interest, depending on the user's selection of the detailed information object (for example, information expressing the degree of overlap between the region of interest and artifacts in more subdivided steps, information indicating the degree of overlap between the region of interest and artifacts numerically, detailed information regarding the portion of the region of interest that overlaps with the artifact, etc.).

[0660] Referring to Figure 37(b), the fourth error information screen may include a fourth notification window CO containing fourth information about the region of interest within the medical image.

[0661] The fourth notification window CO may include a message displaying an error code that includes information indicating that the artifact overlaps with at least part of the area of ​​interest. In this case, the fourth error information screen may include a detailed information object, and the output screen may additionally display detailed information about the fourth information (e.g., an explanation of the error code, the cause of the error message, a manual that can be used to address the error message) based on the user's selection of the detailed information object.

[0662] The fourth notification window CO includes an object that allows the user to choose whether to continue the image analysis if at least part of the region of interest overlaps with the artifact, and the image analysis device 3900 can continue the image analysis according to the user's selection regarding the object for which the image analysis is being performed.

[0663] The fourth notification window CO includes an object that interrupts image analysis if the artifact overlaps with at least a portion of the region of interest, and the image analysis device 3900 may interrupt image analysis depending on...

Claims

1. A device for acquiring brain images and performing morphological analysis based on said brain images is a method for analyzing brain images, The aforementioned method, The method involves obtaining a first correction parameter for correcting a first morphological value associated with a first brain element, wherein the first correction parameter includes a first parameter that corrects the first morphological value taking into account the scan conditions under which the brain image was taken. The method involves obtaining a second correction parameter for correcting a second morphological value associated with a second brain element that is different from a first brain element, wherein the second correction parameter includes a second parameter that corrects the second morphological value considering the scan conditions under which the brain image was taken. Acquiring target brain images and By performing segmentation of the target brain image into a plurality of brain regions including a first region corresponding to the first brain element, a second region corresponding to the second brain element, and a cranial region, a third region related to the internal regions of the first region, the second region, and the cranial region is obtained. Obtaining a first brain morphological index associated with the first brain element based on voxel data corresponding to the first region, voxel data corresponding to the third region, and the first correction parameter, This includes obtaining a second brain morphological index associated with the second brain element based on voxel data corresponding to the second region, voxel data corresponding to the third region, and the second correction parameter, The first morphological value, the second morphological value, the first brain morphological index, and the second brain morphological index are at least one of the volume, length, and thickness of a brain element. Brain imaging analysis methods.

2. Obtaining the first brain morphological indicator or the second morphological indicator is, The method further includes obtaining a reference morphological value related to the internal region based on voxel data corresponding to the third region, Obtaining the aforementioned first neuromorphological indicator is Obtaining a first target morphological value associated with the first brain element based on voxel data corresponding to the first region; calculating a first corrected morphological value associated with the first brain element based on the first target morphological value and the first correction parameter; and calculating the first brain morphological index based on the first corrected morphological value and the reference morphological value. Includes, Obtaining the aforementioned second brain morphological indicator is Obtaining a second target morphological value associated with the second brain element based on voxel data corresponding to the second region; calculating a second corrected morphological value associated with the second brain element based on the second target morphological value and the second correction parameter; and calculating the second brain morphological index based on the second corrected morphological value and the reference morphological value. The brain imaging analysis method according to claim 1, including the method described in claim 1.

3. The first morphological value and the second morphological value are values ​​obtained under the first scan conditions. The target brain image is acquired under the first scan conditions, The first correction parameter includes a first parameter for calculating a third morphological value associated with the first brain element obtained under second scan conditions different from the first scan conditions, based on the first morphological value. The second correction parameter includes a second parameter for calculating a fourth morphological value associated with the second brain element acquired under the second scan conditions based on the second morphological value, Calculating the first corrected morphological value includes obtaining the first corrected morphological value, which is a morphological estimate of the first brain element under the second scan conditions, based on the first target morphological value and the first correction parameter. The brain imaging analysis method according to claim 2, wherein calculating the second corrected morphological value includes obtaining the second corrected morphological value, which is a morphological estimate of the second brain element under the second scan conditions, based on the second target morphological value and the second correction parameter.

4. The first correction parameter includes a first set of parameters associated with a first linear function for calculating the third morphological value based on the first morphological value, The second correction parameter includes a second set of parameters associated with a second linear function for calculating the fourth morphological value based on the second morphological value, Calculating the first corrected morphological value involves obtaining the first corrected morphological value based on the first target morphological value and a first set of parameters associated with the first linear function, The brain imaging analysis method according to claim 3, wherein calculating the second corrected morphological value includes obtaining the second corrected morphological value based on the second target morphological value and a second set of parameters associated with the second linear function.

5. The brain imaging analysis method according to claim 1, wherein the first region corresponding to the first brain element is located adjacent to the skull region, compared to the second region corresponding to the second brain element.

6. The brain imaging analysis method according to claim 1, wherein the first brain element is associated with an element that performs a first brain function, and the second brain element is associated with an element that performs a second brain function different from the first brain function.

7. The brain imaging analysis method according to claim 3, wherein the first scan condition and the second scan condition are related to at least one of the setting parameters related to the resolution of the brain image of the brain imaging device, the magnetic field strength related to the resolution of the brain image of the brain imaging device, the manufacturer of the brain imaging device, and the form of the magnetic field generated by the brain imaging device.

8. The aforementioned target brain image is obtained from a first object having the first characteristic, The first correction parameter and the second correction parameter are obtained from the first brain image and the second brain image obtained from the second object having the first characteristic. The brain imaging analysis method according to claim 1, wherein the first characteristic is related to the age and sex of the subject.

9. The brain imaging analysis method according to claim 1, wherein the segmentation is performed using a neural network provided to acquire a plurality of brain regions corresponding to a plurality of brain elements, including the first brain element, the second brain element, and the skull, based on the target brain imaging.

10. Acquiring the aforementioned target brain image means To acquire initial brain imaging, Preprocessing the aforementioned initial brain images, Includes, Preprocessing the aforementioned initial brain images is This includes performing preprocessing on the first region of the brain image using the first method, and performing preprocessing on the second region of the initial brain image using a second method different from the first method, The brain image analysis method according to claim 4, wherein the segmentation is performed based on the pre-processed target brain image.

11. The first corrected morphological value is a volume value associated with the first brain element acquired in the target brain image, and the second target morphological value is a volume value associated with the second brain element acquired in the target brain image. The first brain morphological index is calculated using the first corrected morphological value relative to the reference morphological value. The brain imaging analysis method according to claim 4, wherein the second brain morphological index is calculated using the second corrected morphological value relative to the reference morphological value.

12. In a device for analyzing brain images, A video acquisition unit that acquires target brain images, A controller that provides brain image analysis information based on the target brain image, The aforementioned controller, The system is configured to acquire a first correction parameter for correcting a first morphological value associated with a first brain element, wherein the first correction parameter includes a first parameter for correcting the first morphological value considering the scan conditions under which the brain image was taken; acquire a second correction parameter for correcting a second morphological value associated with a second brain element different from the first brain element, wherein the second correction parameter includes a second parameter for correcting the second morphological value considering the scan conditions under which the brain image was taken; acquire a target brain image; perform segmentation of the target brain image into a plurality of brain regions including a first region corresponding to the first brain element, a second region corresponding to the second brain element, and a cranial region to acquire a third region associated with the internal regions of the first region, the second region, and the cranial region; acquire a first brain morphological index associated with the first brain element based on voxel data corresponding to the first region, voxel data corresponding to the third region, and the first correction parameter; and acquire a second brain morphological index associated with the second brain element based on voxel data corresponding to the second region, voxel data corresponding to the third region, and the second correction parameter. A brain imaging analyzer wherein the first morphological value, the second morphological value, the first brain morphological index, and the second brain morphological index are associated with at least one of the volume, length, and thickness of a brain element.

13. The aforementioned controller, The system is configured to acquire reference morphological values ​​related to the internal region based on voxel data corresponding to the third region, Based on the voxel data corresponding to the first region, a first target morphological value associated with the first brain element is obtained; based on the first target morphological value and the first correction parameter, a first corrected morphological value associated with the first brain element is calculated; based on the first corrected morphological value and the reference morphological value, the first brain morphological index is calculated and the first brain morphological index is obtained. The brain image analysis device according to claim 12, configured to acquire a second target morphological value associated with the second brain element based on voxel data corresponding to the second region, calculate a second corrected morphological value associated with the second brain element based on the second target morphological value and the second correction parameter, and calculate the second brain morphological index based on the second corrected morphological value and the reference morphological value to acquire the second brain morphological index.

14. The first morphological value and the second morphological value are values ​​obtained under the first scan conditions. The target brain image is acquired under the first scan conditions, The first correction parameter includes a first parameter for calculating a third morphological value associated with the first brain element obtained under second scan conditions different from the first scan conditions, based on the first morphological value. The second correction parameter includes a second parameter for calculating a fourth morphological value associated with the second brain element acquired under the second scan conditions based on the second morphological value, The aforementioned controller, The brain imaging analyzer according to claim 13, configured to calculate a first corrected morphological value, which is a morphological estimate of the first brain element under the second scan conditions, based on the first target morphological value and the first correction parameter, and to calculate a second corrected morphological value, which is a morphological estimate of the second brain element under the second scan conditions, based on the second target morphological value and the second correction parameter.

15. The first correction parameter includes a first set of parameters associated with a first linear function for calculating the third morphological value based on the first morphological value, The second correction parameter includes a second set of parameters associated with a second linear function for calculating the fourth morphological value based on the second morphological value, Calculating the first corrected morphological value involves obtaining the first corrected morphological value based on the first target morphological value and a first set of parameters associated with the first linear function, The aforementioned controller, The brain imaging analysis device according to claim 14, configured to calculate a first corrected morphological value based on a first set of parameters associated with the first target morphological value and the first linear function, and to calculate a second corrected morphological value based on a second set of parameters associated with the second target morphological value and the second linear function.

16. The brain imaging analyzer according to claim 12, wherein the first region corresponding to the first brain element is located adjacent to the skull region, compared to the second region corresponding to the second brain element.

17. The video analysis apparatus according to claim 12, wherein the first brain element is associated with an element that performs a first brain function, and the second brain element is associated with an element that performs a second brain function different from the first brain function.

18. The video analysis apparatus according to claim 14, wherein the first scan condition and the second scan condition are associated with at least one of the setting parameters associated with the resolution of the brain image acquired by

19. The aforementioned controller, The aforementioned target brain image is obtained from a first object having the first characteristic, The first correction parameter and the second correction parameter are obtained from the first brain image and the second brain image obtained from the second object having the first characteristic. The image analysis apparatus according to claim 12, wherein the first characteristic relates to the age and gender of the subject.

20. The video analysis device according to claim 12, wherein the segmentation is performed using a neural network provided to acquire a plurality of brain regions corresponding to a plurality of brain elements, including the first brain element, the second brain element, and the skull, based on the target brain image.

21. The aforementioned controller, The video analysis apparatus according to claim 15, wherein the target brain image is preprocessed, the segmentation is performed based on the preprocessed target brain image, and the preprocessing is performed on the first region using a first method and on the second region using a preprocessing method different from the first method.

22. The first corrected morphological value is a volume value associated with the first brain element acquired in the target brain image, and the second target morphological value is a volume value associated with the second brain element acquired in the target brain image. The video analysis apparatus according to claim 15, wherein the first brain morphological index is calculated using the first corrected morphological value relative to the reference morphological value, and the second brain morphological index is calculated using the second corrected morphological value relative to the reference morphological value.

Images