A medical image digital assisted analysis method and system

Abstract

The present application discloses a method and system for auxiliary analysis of medical images, relating to data processing and image processing technologies, including: pre-establishing a typical three-dimensional model containing a target organ and acquiring a medical image containing an organ lesion area; aligning the medical image on corresponding sections based on the typical three-dimensional model, and identifying the lesion area in the medical image; mapping the lesion area based on the aligned sections of the typical three-dimensional model; determining at least one local center point from the lesion area to perform local scanning on the typical three-dimensional model based on the local center point and the lesion boundary to establish a three-dimensional range of the local lesion; and presenting to the user based on the typical three-dimensional model after establishing the three-dimensional range of the local lesion. The method of the present application can perform auxiliary section according to user operations and improve the auxiliary effect of medical images.

Description

Technical Field

[0001] This application relates to the technical fields of data processing and image processing, and particularly to a method and system for digital assisted analysis of medical images. Background Art

[0002] Computer-aided analysis of medical images uses advanced computer software and hardware systems to analyze and process digital radiological images to discover and detect lesion features. The results are used as a "second opinion" for reference by diagnostic physicians, helping radiologists improve the detection rate of lesions, improving diagnostic accuracy and the reproducibility of diagnosis, shortening the film reading time, and improving work efficiency.

[0003] In the prior art, traditional computer-aided measurement (CAM) and computer-aided diagnosis (CAD) technologies used in medical ultrasound imaging systems have many problems and limitations. Their detection results are one-sided and non-intuitive, and they cannot truly achieve a comprehensive and accurate analysis of the organ conditions. Summary of the Invention

[0004] Embodiments of this application provide a method and system for digital assisted analysis of medical images, which are used to determine the scope of the lesion area by combining the three-dimensional model of a typical organ and the medical image of a patient, and perform an auxiliary section according to user operations to improve the auxiliary effect of medical images.

[0005] Embodiments of this application propose a method for digital assisted analysis of medical images, including:

[0006] Pre-establish a typical three-dimensional model containing the target organ, and obtain a medical image containing the organ lesion area;

[0007] Align the medical image on the corresponding section based on the typical three-dimensional model, and identify the lesion area in the medical image;

[0008] Map the lesion area based on the aligned section of the typical three-dimensional model;

[0009] Determine at least one local center point from the lesion area, and perform a local scan on the typical three-dimensional model based on the local center point and the lesion boundary to establish a three-dimensional range of the local lesion;

[0010] Based on the typical three-dimensional model after establishing the three-dimensional range of the local lesion, present it to the user, and according to the selected rotation angle of the user, present a section image containing the local lesion area in association with the typical three-dimensional model.

[0011] Optionally, aligning the medical image on the corresponding section based on the typical three-dimensional model includes:

[0012] Extract the organ boundary in the medical image, and determine multiple reference positions from the medical image outside the organ boundary;

[0013] Match the organ boundary with the typical three-dimensional model to determine an alignment interval in the typical three-dimensional model;

[0014] Perform a section in the alignment interval to obtain a sectional image, and use each sectional image to perform a secondary match with the organ boundary and each reference position respectively, so as to use the section with the maximum determined matching degree as the aligned sectional image to complete the alignment.

[0015] Optionally, performing a slice in the alignment interval to obtain a slice image, and using each slice image to perform a secondary match with the organ boundary and each reference position respectively includes:

[0016] In the alignment interval area of the typical three-dimensional model, determine a first sectional image by sectioning at an arbitrary angle;

[0017] Based on the first sectional image, calculate the specification relationship at each reference position in the medical image;

[0018] Adjust the sectioning angle according to the change trend of the specification relationship along the organ boundary;

[0019] Repeat sectioning the typical three-dimensional model according to the adjusted sectioning angle to obtain a second sectional image, and repeat calculating the specification relationship at each reference position in the medical image until the calculated specification relationships are all within a preset fluctuation interval.

[0020] Optionally, adjusting the sectioning angle according to the change trend of the specification relationship along the organ boundary includes:

[0021] Determine the reference position where the specification relationship along the organ boundary is closest;

[0022] Determine the specification change trends of other reference positions on both sides of the reference position where the specification relationship is closest;

[0023] Determine the required adjusted sectioning angle according to the increase and decrease relationship of the specification change trends of other reference positions.

[0024] Optionally, determine at least one local center point from the lesion area, and perform a local scan on the typical three-dimensional model based on the local center point and the lesion boundary to establish a local lesion three-dimensional range, including:

[0025] According to the identified lesion area in the medical image, determine multiple arc segments and determine the center of each arc segment;

[0026] Cluster the determined centers to obtain at least one local center point;

[0027] Based on the local center point and the lesion boundary, perform local scanning on the typical three-dimensional model based on the mapped lesion area to establish the three-dimensional range of the local lesion.

[0028] Optionally, based on the local center point and the lesion boundary, performing local scanning on the typical three-dimensional model based on the mapped lesion area includes:

[0029] Perform surface scanning on the typical three-dimensional model respectively based on each local center point and the lesion boundary;

[0030] Determine the connection boundaries of the surface scanning results;

[0031] Retain one of the regions with overlapping parts in any two surface scanning results, and retain the regions outside each connection boundary;

[0032] Smooth each connection boundary to establish the three-dimensional range of the local lesion.

[0033] Optionally, smoothing each connection boundary includes: for the concave boundary in the connection boundary, take the regions on both sides of the concave boundary within a preset pixel distance as the smoothed boundary.

[0034] Optionally, presenting to the user based on the typical three-dimensional model after establishing the three-dimensional range of the local lesion includes:

[0035] Judge the rotation angle of the user for the typical three-dimensional model after establishing the three-dimensional range of the local lesion;

[0036] According to the rotation angle and the model display direction of the screen, perform a section on the typical three-dimensional model after establishing the three-dimensional range of the local lesion at the largest region of the three-dimensional range of the local lesion to present to the user.

[0037] An embodiment of the present application also proposes a medical image digital auxiliary analysis system, including a processor and a memory, and a computer program is stored on the memory. When the computer program is executed by the processor, the steps of the medical image digital auxiliary analysis method as described above are implemented.

[0038] The method of the embodiment of the present application combines the three-dimensional model of the typical organ and the medical image of the patient to determine the range of the lesion area, and performs an auxiliary section according to the user operation, which can effectively improve the auxiliary effect of the medical image on the doctor.

[0039] The above description is only an overview of the technical solution of the present application. In order to be able to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the specific embodiments of the present invention are specifically exemplified below. Brief Description of the Drawings

[0040] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0041] Figure 1 It is a schematic diagram of the basic process of the medical image digital assistant analysis method of this embodiment. Detailed Embodiments

[0042] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art.

[0043] An embodiment of the present application proposes a medical image digital assistant analysis method, as Figure 1 shown, including the following steps:

[0044] In step S101, a typical three-dimensional model including the target organ is established in advance, and a medical image including the lesion area of the organ is obtained. In some embodiments, such as organs like the heart and liver, typical three-dimensional models corresponding to the respective age groups of the patient can be established. In some specific examples, a typical three-dimensional model including the organ part and the nearby tissues can be established.

[0045] In step S102, the medical image is aligned on the corresponding section based on the typical three-dimensional model, and the lesion area in the medical image is identified. In a specific example, for the imaging of an organ or a lesion, a medical image with the largest contour range can generally be obtained, and based on this, the section can be aligned to the typical three-dimensional model.

[0046] In step S103, based on the aligned section of the typical three-dimensional model, the lesion area is mapped. That is, the lesion area is mapped according to the aligned typical three-dimensional model.

[0047] In step S104, at least one local center point is determined from the lesion area, and local scanning is performed on the typical three-dimensional model based on the local center point and the lesion boundary to establish a local lesion three-dimensional range.

[0048] In step S105, based on the typical three-dimensional model after establishing the three-dimensional range of the local lesion, it is presented to the user, and according to the rotation angle selected by the user, a sectional image including the local lesion area is presented in association with the typical three-dimensional model. In a specific example, the three-dimensional model at a set angle can be default presented according to the type of the organ, and the lesion area is sectioned at this angle, so that, for example, the three-dimensional model and the sectional image are presented on the interface at the same time.

[0049] The method of the embodiment of the present application combines the three-dimensional model of the typical organ and the medical image of the patient to determine the range of the lesion area, and performs an auxiliary section according to the user operation, which can effectively improve the auxiliary effect of the medical image on the doctor.

[0050] In some embodiments, aligning the medical image on the corresponding section based on the typical three-dimensional model includes:

[0051] Extract the organ boundary in the medical image, and determine a plurality of reference positions from the medical image outside the organ boundary. For example, for the medical image of the lung, a plurality of reference positions can be determined according to the lung boundary and the landmark tissues or structures outside the boundary.

[0052] Match the organ boundary with the typical three-dimensional model to determine an alignment interval in the typical three-dimensional model.

[0053] Perform a section in the alignment interval to obtain a sectional image, and use each sectional image to perform a secondary match with the organ boundary and each reference position respectively, so as to use the sectional image with the largest determined matching degree as the aligned sectional image to complete the alignment. That is, by determining the alignment interval in the typical three-dimensional model and performing an accurate secondary match according to a plurality of reference positions in the alignment interval, the medical image is matched to the typical three-dimensional model of the target organ established.

[0054] In some embodiments, performing a slice in the alignment interval to obtain a slice image and using each slice image to perform a secondary match with the organ boundary and each reference position respectively includes:

[0055] In the area of the alignment interval in the typical three-dimensional model, a first sectional image is determined by sectioning at an arbitrary angle. This example is for the fine alignment step to further perform in the area of the alignment interval determined by the foregoing rough alignment. The area of the alignment interval is sectioned in the typical three-dimensional model to obtain the first sectional image.

[0056] Based on the first sectional image, calculate the specification relationships at each reference position in the medical image. The specification relationships referred to in the embodiments of the present application can be size relationships. For example, taking the extracted medical image as a reference, if the specification of the reference position in the first sectional image is smaller than that of the medical image, the size relationship is negative; if the specification of the reference position in the first sectional image is larger than that of the medical image, the size relationship is positive; and if they exactly match, it approaches 0.

[0057] Adjust the sectional angle according to the changing trend of the specification relationship along the organ boundary, and use the changing trend of the size relationship outside the boundary described above as a reference to adjust the sectional angle.

[0058] Repeat obtaining the second sectional image for the typical three-dimensional model section according to the adjusted sectional angle, and repeat calculating the specification relationships at each reference position in the medical image until the calculated specification relationships are all within a preset fluctuation range. Repeating the above operations can complete the fine alignment step.

[0059] In some embodiments, adjusting the sectional angle according to the changing trend of the specification relationship along the organ boundary includes:

[0060] Determine the reference position along the organ boundary where the specification relationship is closest to the reference, that is, the reference position closest to 0.

[0061] Determine the changing trends of the specifications of other reference positions on both sides of the reference position where the specification relationship is closest. According to the other changing trends on both sides of the reference position, such as the positive and negative gradual increase of the specification relationships on both sides, the sectional angle can be finely adjusted according to this increase and decrease relationship, so as to complete the fine alignment.

[0062] Determine the required adjusted sectional angle according to the increase and decrease relationship of the changing trends of the specifications of other reference positions.

[0063] In some embodiments, determining at least one local center point from the lesion area to perform local scanning on the typical three-dimensional model based on the local center point and the lesion boundary to establish a local lesion three-dimensional range includes:

[0064] According to the identified lesion area in the medical image, determine a plurality of arc segments and determine the centers of each arc segment. In some embodiments, the arc segments in the lesion area of the medical image can be divided, and then the centers corresponding to the plurality of arc segments can be determined.

[0065] Cluster the determined centers of each arc segment to obtain at least one local center point. For example, if the centers determined for a plurality of small local arc segments are approximately in the same position, they are clustered into one center point.

[0066] Based on the local center point and the lesion boundary, in the typical three-dimensional model, local scanning is performed based on the mapped lesion area to establish the three-dimensional range of the local lesion. The scanning referred to in the embodiments of the present application can be performed within the typical three-dimensional model on a spherical surface with the local center point as the center of the sphere and the corresponding arc segment as the radius. In a specific example, medical images of the target organ at multiple angles can be used to determine the three-dimensional range of the accurate lesion. Further, for the lines outside the determined arc segment, such as straight lines and quasi-straight lines, direct mapping is performed, and then the scanning results of medical images at multiple angles can be stitched and combined in the lesion area to determine the three-dimensional range of the lesion in the typical three-dimensional model.

[0067] In some embodiments, based on the local center point and the lesion boundary, performing local scanning in the typical three-dimensional model based on the mapped lesion area includes:

[0068] Performing surface scanning separately in the typical three-dimensional model based on each local center point and the lesion boundary;

[0069] Determining the connection boundaries of the results of each surface scan.

[0070] Retaining one of the regions with an overlapping part in any two surface scan results, and retaining the regions outside each connection boundary, that is, taking the union in the scan results as the three-dimensional range of the local lesion.

[0071] Smoothing each connection boundary to establish the three-dimensional range of the local lesion.

[0072] In some embodiments, smoothing each connection boundary includes: for the concave boundary in the connection boundary, taking the regions on both sides of the concave boundary with a distance within the preset pixel distance as the smoothed boundary. In a specific example, the lesion body region formed after the combination of the connection boundaries is not smooth. The shortest pixel distance between the two sides of the determined boundary is determined. If the shortest pixel distance is less than the preset pixel distance, corresponding pixel filling is performed to complete the boundary smoothing.

[0073] In some embodiments, presenting to the user based on the typical three-dimensional model after establishing the three-dimensional range of the local lesion includes:

[0074] Judging the rotation angle of the user with respect to the typical three-dimensional model after establishing the three-dimensional range of the local lesion;

[0075] According to the rotation angle and the model display direction of the screen, performing a section on the largest region of the three-dimensional range of the local lesion of the typical three-dimensional model after establishing the three-dimensional range of the local lesion to present to the user.

[0076] In a specific example, by cutting a section at the largest area within the three-dimensional range of a local lesion on a typical three-dimensional model after establishing the three-dimensional range of the local lesion, it is possible to assist doctors in observing and analyzing the lesion area from various angles, thereby improving the efficiency of digital assisted analysis of medical images.

[0077] An embodiment of the present application also proposes a medical image digital assisted analysis system, including a processor and a memory. A computer program is stored on the memory, and when the computer program is executed by the processor, the steps of the medical image digital assisted analysis method as described above are implemented.

[0078] In addition, although exemplary embodiments have been described herein, the scope includes any and all embodiments based on the present disclosure having equivalent elements, modifications, omissions, combinations (e.g., solutions that cross various embodiments), adaptations, or alterations. It is not limited to the examples described in this specification or during the implementation of this application, and the examples will be construed as non-exclusive.

[0079] The above description is intended to be illustrative rather than restrictive. For example, the above examples (or one or more of their solutions) can be used in combination with each other. For example, those of ordinary skill in the art can use other embodiments when reading the above description.

[0080] The above embodiments are only exemplary embodiments of the present disclosure. Those skilled in the art can make various modifications or equivalent replacements to the present invention within the essence and protection scope of the present disclosure, and such modifications or equivalent replacements should also be regarded as falling within the protection scope of the present invention.

Claims

1. A method for digital assisted analysis of medical images, characterized in that Comprising: Pre - establish a pre - stored standardized typical three - dimensional model including the target organ, and obtain a medical image containing the organ lesion area; Extract the organ boundary from the medical image, and determine a plurality of reference positions from the medical image outside the organ boundary; Match the organ boundary with the typical three - dimensional model to determine an alignment interval in the typical three - dimensional model; Perform a section in the alignment interval to obtain a section image, and use each section image to perform a secondary match with the organ boundary and each reference position based on gradient similarity, so as to take the section with the maximum determined matching degree as the aligned section image, complete the alignment on the corresponding section, and identify the lesion area in the medical image, where in the secondary match, determine that the specification relationship along the organ boundary is closest to the reference position; Determine the specification change trend of other reference positions on both sides of the reference position with the closest specification relationship; determine the section angle to be adjusted according to the increase - decrease relationship of the specification change trends of other reference positions; Map the lesion area based on the aligned section in the typical three - dimensional model; Determine a plurality of arc segments according to the identified lesion area in the medical image, and determine the center of each arc segment; Cluster the determined centers to obtain at least one local center point; Based on the local center point and the lesion boundary, perform a local scan on the mapped lesion area in the typical three - dimensional model to establish a local lesion three - dimensional range; Based on the typical three - dimensional model after establishing the local lesion three - dimensional range, present it to the user, and according to the selected rotation angle of the user, dynamically associate and present a section image containing the local lesion area with the typical three - dimensional model.

2. The medical image digital assisted analysis method according to claim 1, wherein Performing slicing in the alignment interval to obtain a slice image, and using each slice image to perform a secondary match with the organ boundary and each reference position respectively includes: In the alignment interval area of the typical three - dimensional model, determine a first section image by slicing at an arbitrary angle; Based on the first section image, calculate the specification relationship at each reference position in the medical image; Adjust the section angle according to the change trend of the specification relationship along the organ boundary; Repeat slicing the typical three - dimensional model according to the adjusted section angle to obtain a second section image, and repeat calculating the specification relationship at each reference position in the medical image until the calculated specification relationships are all within a preset fluctuation interval.

3. The medical image digital assisted analysis method according to claim 1, wherein, Based on the local center point and the lesion boundary, performing a local scan on the mapped lesion area in the typical three - dimensional model includes: Based on each local center point and the lesion boundary, perform surface scans in the typical three - dimensional model respectively; Determine the connection boundaries of the results of each surface scan; Retain one of the regions with overlapping parts in any two surface scan results, and retain the regions outside each connection boundary; Smooth each connection boundary to establish a local lesion three - dimensional range.

4. The medical image digital assisted analysis method according to claim 3, characterized in that, Smoothing each connection boundary includes: for the concave boundary in the connection boundary, take the region where the distance between both sides of the concave boundary is within a preset pixel distance as the smoothed boundary.

5. The medical image digital assisted analysis method according to claim 1, characterized in that, Presenting to the user based on the typical three - dimensional model after establishing the local lesion three - dimensional range includes: Determine the rotation angle of the typical three-dimensional model after establishing the three-dimensional range of the local lesion by the user; According to the rotation angle and the model display direction of the screen, perform a section on the maximum area of the three-dimensional range of the local lesion of the typical three-dimensional model after establishing the three-dimensional range of the local lesion, so as to present it to the user.

6. A medical image digital auxiliary analysis system, characterized in that, It includes a processor and a memory, and a computer program is stored on the memory. When the computer program is executed by the processor, the steps of the medical image digital auxiliary analysis method described in any one of claims 1 to 5 are implemented.

Images