Medical image processing device, medical image processing method, and medical image processing program

Abstract

To improve accuracy of quantitative evaluation for mismatching of two medical images.SOLUTION: A medical image processing device includes: an acquisition unit for acquiring a first medical image having an imaging region and a second medical image collected by the imaging different from the imaging on the first medical image and including the imaging region; a specification unit for specifying a first region on a lesion in the first medical image; a determination unit for determining whether the first region is bilateral or unilateral with respect to the imaging region; a decision unit for deciding a region different from the first region as a second region indicating a normal part in the second medical image if the first region is bilateral, and deciding a region corresponding to the line symmetry of the first region with respect to a center line of the imaging region as the second region in the second medical image if the first region is unilateral; and a calculation unit for calculating a first feature amount on the first region and a second feature amount on the second region in the second medical image, and calculating an index on mismatching of the first and second medical images on the basis of the first and second feature amounts.SELECTED DRAWING: Figure 6

Description

Technical Field

[0001] The embodiments disclosed in this specification and the drawings relate to a medical image processing apparatus, a medical image processing method, and a medical image processing program.

Background Art

[0002] Conventionally, as a method for predicting the onset time of cerebral infarction, there is a method that focuses on the mismatch (DWI / FLAIR mismatch phenomenon) between a diffusion-weighted image (DWI) and a FLAIR (Fluid Attenuated Inversion Recovery) image among magnetic resonance imaging (MRI) images.

[0003] A DWI image can depict acute cerebral infarction. On the other hand, in acute cerebral infarction, no change is seen in the FLAIR image. For these reasons, as shown in FIG. 10, the user uses the difference (mismatch) between the DWI image and the FLAIR image to determine whether the cerebral infarction of the subject is acute cerebral infarction. As shown in FIG. 10, when there is a mismatch (TMM) between the DWI image and the FLAIR image, it corresponds to acute cerebral infarction, and when there is no mismatch (NMM), it often corresponds to non-acute cerebral infarction.

[0004] In the case of acute cerebral infarction, for example, the effect of thrombus removal by administration of rt-PA (recombinant tissue-type plasminogen activator) is recognized. For this reason, there is a desire to grasp whether the cerebral infarction that has occurred in the subject is acute. Cerebral infarction often occurs from midnight to early morning, such as during sleep, and the onset time of cerebral infarction is often unknown. For this reason, by focusing on the DWI / FLAIR mismatch phenomenon, it may be possible to grasp the approximate elapsed time since the onset of cerebral infarction.

[0005] In the DWI / FLAIR mismatch phenomenon, first, the area related to cerebral infarction in the DWI image is identified as the area on the affected side (hereinafter referred to as the affected area), and the affected area is mapped to the FLAIR image. Next, in the FLAIR image, the area corresponding to the line symmetry of the affected area with the center line of the brain as the axis of symmetry is determined as the area on the healthy side (hereinafter referred to as the healthy area). Subsequently, it is known that the DWI / FLAIR mismatch phenomenon is quantified using the affected area and the healthy area in the FLAIR image.

[0006] However, when there are infarctions on both the left and right sides of the subject's brain, such as cerebellar infarction or bilateral infarction, in the quantification of the known DWI / FLAIR mismatch phenomenon, the affected side and the healthy side cannot be distinguished. For this reason, it becomes difficult to quantitatively evaluate the DWI / FLAIR mismatch phenomenon using the affected area and the healthy area. Therefore, the doctor is qualitatively judging the DWI / FLAIR mismatch phenomenon in his own mind.

Prior Art Documents

Non-Patent Documents

[0007]

Non-Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0008] One of the problems to be solved by the embodiments disclosed in this specification and the drawings is to improve the accuracy of quantitative evaluation of mismatches in two medical images regardless of whether the lesions are bilateral or unilateral. However, the problems to be solved by the embodiments disclosed in this specification and the drawings are not limited to the above problems. The problems corresponding to the effects of each configuration shown in the embodiments described later can also be regarded as other problems.

Means for Solving the Problems

[0009] The medical image processing apparatus according to this embodiment includes an acquisition unit, a specification unit, a determination unit, a determination unit, and a calculation unit. The acquisition unit acquires a first medical image collected by a predetermined imaging of an imaging site of a subject, and a second medical image collected by an imaging different from the predetermined imaging and including the imaging site. The specification unit specifies a first region related to a lesion in the imaging site based on the first medical image. The determination unit determines whether the first region is bilateral or unilateral with respect to the imaging site based on the position of the first region in the imaging site. When it is determined that the first region is bilateral, the determination unit determines a region different from the first region in the second medical image as a second region indicating a normal part in the imaging site. When it is determined that the first region is unilateral, the determination unit determines a region corresponding to the line symmetry of the first region with the center line of the imaging site in the second medical image as the second region. The calculation unit calculates a first feature amount indicating a feature by a plurality of pixel values in the first region and a second feature amount indicating a feature by a plurality of pixel values in the second region in the second medical image, and calculates an index related to the mismatch between the first medical image and the second medical image based on the first feature amount and the second feature amount.

Brief Description of the Drawings

[0010]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

[0011] Hereinafter, embodiments of a medical image processing apparatus, a medical image processing method, and a medical image processing program will be described with reference to the drawings. In the following embodiments, parts denoted by the same reference numerals perform the same operations, and redundant descriptions will be omitted as appropriate.

[0012] (Embodiment) FIG. 1 is a block diagram showing an example of the configuration of a medical information processing system 1 including a medical image processing apparatus 30 according to an embodiment. As shown in FIG. 1, the medical information processing system 1 according to the embodiment includes a magnetic resonance imaging (MRI) apparatus 10, an image storage apparatus 20, and a medical image processing apparatus 30. As shown in FIG. 1, the MRI apparatus 10, the image storage apparatus 20, and the medical image processing apparatus 30 are interconnected via a network.

[0013] The MRI apparatus 10 collects magnetic resonance images (MR images) from a subject P. For example, the MRI apparatus 10 generates an MR image by collecting and reconstructing MR data from the subject P. The MRI apparatus 10 transmits the generated MR image to the image storage apparatus 20 or the medical image processing apparatus 30. Since a known configuration can be applied as the configuration of the MRI apparatus 10, the description thereof is omitted. The MRI apparatus 10 is an example of an imaging apparatus.

[0014] The MRI images shall be a first medical image and a second medical image. The first medical image is collected, for example, by a predetermined imaging of an imaging site of the subject. Hereinafter, in order to specifically describe, the first medical image shall be a diffusion weighted imaging (DWI). The diffusion weighted imaging also includes an ADC (apparent diffusion coefficient) map. Further, the first medical image shall be an axial cross-section of the imaging site or a coronal cross-section of the imaging site. Further, the imaging site shall be the brain of the subject. The predetermined imaging shall be, for example, diffusion weighted (DW) imaging (DW-EPI) using echo planar imaging (EPI). Since a known imaging method of a T2 weighted image system can be applied as the DW imaging, the description thereof is omitted. The DW imaging is typically performed on an axial cross-section of the imaging site or a coronal cross-section of the imaging site.

[0015] The second medical image is collected by an imaging different from the predetermined imaging and includes the imaging site related to the collection of the first medical image. Assume that the second medical image is a FLAIR (Fluid Attenuated Inversion Recovery) image. Also, assume that the second medical image is an axial cross-section of the imaging site or a coronal cross-section of the imaging site. Also, assume that the imaging site is the brain of the subject, similar to the first medical image. The imaging different from the predetermined imaging is, for example, FLAIR imaging (FLAIR-FSE) using the Fast Spin echo method. Since known imaging techniques of the water suppression T2-weighted image system are applicable as FLAIR imaging, the description is omitted. FLAIR imaging is typically performed on an axial cross-section of the imaging site or a coronal cross-section of the imaging site. Note that the second medical image may be the same cross-section as the first medical image or a different cross-section from the first medical image.

[0016] Note that DWI imaging and FLAIR imaging are not limited to an axial cross-section of the imaging site or a coronal cross-section of the imaging site, and may be performed on a volume scan of the subject's brain or other cross-sections such as an oblique cross-section. At this time, the first medical image and the second medical image are generated by a cross-section conversion process on the generated volume data.

[0017] The image storage device 20 stores the first medical image and the second medical image collected by the MRI device 10, etc. For example, the image storage device 20 is realized by a computer device such as a server device. Specifically, the image storage device 20 is realized by a PACS (Picture Archiving and Communication System) server or the like. The image storage device 20 may be referred to as a medical image management system. In the present embodiment, the image storage device 20 acquires the first medical image and the second medical image from the MRI device 10 via a network, and stores the acquired first medical image and second medical image in a memory provided inside or outside the device.

[0018] The medical image processing device 30 acquires the first medical image and the second medical image from the MRI device 10 or the image storage device 20 via a network, and executes various processes using the acquired first medical image and second medical image. The medical image processing device 30 may also be referred to as a medical image analysis device or an analysis device. For example, the medical image processing device 30 is realized by a computer device such as a workstation. Further, the medical image processing device 30 causes the display 32 to display the result of the process processed based on the first medical image and the second medical image.

[0019] As shown in FIG. 1, the medical image processing device 30 includes an input interface 31, a display 32, a memory 33, and a processing circuit 34.

[0020] The input interface 31 is realized by a trackball, a switch, a button, a mouse, a keyboard, a touch pad that performs an input operation by touching an operation surface, a touch screen in which a display screen and a touch pad are integrated, a non-contact input circuit using an optical sensor, a voice input circuit, etc. for performing various instructions and various settings. The input interface 31 converts the input operation received from the operator into an electrical signal and outputs it to the processing circuit 34. Note that the input interface 31 is not limited to those provided with physical operation components such as a mouse and a keyboard. For example, an electrical signal processing circuit that receives an electrical signal corresponding to an input operation from an external input device provided separately from the medical image processing device 30 and outputs this electrical signal to the processing circuit 34 is also included in the example of the input interface 31. The input interface 31 is an example of an input unit.

[0021] The display 32 displays various information under the control of the display control function 34f. For example, the display 32 displays a GUI (Graphical User Interface) for receiving an operator's instruction and various X-ray image data. For example, the display 32 is a liquid crystal display or a CRT (Cathode Ray Tube) display. The display 32 is an example of a display unit.

[0022] The memory 33 is realized by, for example, a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory, a hard disk, an optical disk, or the like. For example, the memory 33 stores the first medical image and the second medical image acquired from the MRI apparatus 10 or the image storage apparatus 20. Further, for example, the memory 33 stores a program for each circuit included in the medical image processing apparatus 30 to realize its function. The memory 33 is an example of a storage unit.

[0023] The processing circuit 34 controls the operation of the entire medical image processing apparatus 30 by executing an acquisition function 34a, a specification function 34b, a determination function 34c, a decision function 34d, a calculation function 34e, and a display control function 34f.

[0024] The processing circuit 34 reads out and executes a program corresponding to the acquisition function 34a from the memory 33, thereby acquiring the first medical image and the second medical image from the MRI apparatus 10 or the image storage apparatus 20. The acquisition function 34a stores the first medical image and the second medical image in the memory 33. The processing circuit 34 that realizes the acquisition function 34a corresponds to an acquisition unit.

[0025] The processing circuit 34 reads out and executes a program corresponding to the specification function 34b from the memory 33. As a result, the specification function 34b specifies a first region related to a lesion in the imaging region based on the first medical image. For example, the specification function 34b executes image processing on the first medical image and specifies a first region in the first medical image. For example, the specification function 34b specifies, in the ADC map in the first medical image, a plurality of pixels having a pixel value lower than a predetermined threshold as the first region. The predetermined threshold corresponds to, for example, a value at which the apparent diffusion coefficient (ADC) is lower than that of a healthy part (hereinafter referred to as the ADC threshold). The ADC threshold is preset and stored in the memory 33. A small value of the apparent diffusion coefficient corresponds to a tissue affected by a vascular infarction. Therefore, when the imaging region is the brain, the first region corresponds to the region where a cerebral infarction has occurred.

[0026] Note that the identification of the first region is not limited to the segmentation process using the ADC threshold value for the ADC map in the first medical image. For example, a trained model that takes the first medical image as input and outputs the first region may be used for the identification of the first region.

[0027] Further, the processing circuit 34, by the specific function 34b, identifies the positional association between the first medical image and the second medical image by registering the first medical image and the second medical image. The positional association corresponds to, for example, a transformation matrix (mapping matrix) for associating each of a plurality of pixels in the first medical image with each of a plurality of pixels in the second medical image. Since known methods are applicable for the registration and the generation of the mapping matrix, the description thereof is omitted. The specific function 34b stores the mapping matrix in the memory 33.

[0028] Further, the processing circuit 34, by the specific function 34b, identifies the center line with respect to the imaging site in the second medical image. When the imaging site is a substantially left-right symmetric organ such as the brain (i.e., a substantially line-symmetric organ), the center line corresponds to the axis of symmetry with respect to the approximate left-right symmetry, that is, the axis of symmetry with respect to the substantially line-symmetric organ. Since known methods are applicable for the image processing related to the identification of the center line, the description thereof is omitted. The specific function 34b stores the center line (axis of symmetry) identified with respect to the imaging site in the second medical image in the memory 33. The processing circuit 34 that realizes the specific function 34b corresponds to the specific unit.

[0029] The processing circuit 34, by the determination function 34c, determines whether the first region is bilateral or unilateral with respect to the imaging site based on the position of the first region in the imaging site. The position of the first region in the imaging site corresponds to, for example, the position of the first region with respect to the axis of symmetry in the left-right symmetric imaging site. The determination function 34c uses the position of the first region in the imaging site in the second medical image to determine whether the first region is bilateral or unilateral with respect to the imaging site. The processing circuit 34 that realizes the determination function 34c corresponds to the determination unit.

[0030] Specifically, the processing circuit 34 uses the determination function 34c to map the first region to the second medical image using the mapping matrix. Next, the determination function 34c calculates the line symmetry of the first region in the second medical image with the center line of the imaging site in the second medical image as the axis of symmetry, and identifies the symmetric region corresponding to the line symmetry. The determination function 34c determines whether the first region and the symmetric region overlap in the second medical image. If the first region and the symmetric region overlap, the determination function 34c determines that the first region is bilateral. Also, if the first region and the symmetric region do not overlap, the determination function 34c determines that the first region is unilateral. In the above description, the bilateral or unilateral nature of the first region is determined in the second medical image, but it is not limited to this, and the bilateral or unilateral nature of the first region may be determined in the first medical image.

[0031] FIG. 2 is a diagram showing an example of the first region 1RE and the symmetric region SR in the second medical image 2ME. As shown in FIG. 2, the symmetric region SR is set at a position corresponding to the line symmetry of the first region 1RE with the center line CL as the axis of symmetry. As shown in FIG. 2, since the first region 1RE and the symmetric region SR do not overlap, the first region 1RE is determined to be unilateral by the determination function 34c.

[0032] FIG. 3 is a diagram showing an example of the first region 1RE and the symmetric region SR in the second medical image 2ME. As shown in FIG. 3, a part of the symmetric region SR set at a position corresponding to the line symmetry of the first region 1RE with the center line CL as the axis of symmetry overlaps with the first region 1RE. For this reason, the first region 1RE is determined to be bilateral by the determination function 34c.

[0033] Note that when the size of the overlapping region where the symmetric region SR and the first region 1RE overlap exceeds a predetermined threshold by the determination function 34c, the processing circuit 34 may determine that the first region 1RE is bilateral, and when the size of the overlapping region is equal to or less than the predetermined threshold, the processing circuit 34 may determine that the first region 1RE is unilateral. The predetermined threshold is a threshold related to the determination of bilateral or unilateral (hereinafter referred to as the determination threshold), and is stored in a memory 33 preset with, for example, volume, area, number of voxels, number of pixels, etc.

[0034] FIG. 4 is a diagram showing an example of the first region 1RE and the symmetric region SR in the second medical image 2ME. As shown in FIG. 4, the symmetric region SR is set at a position corresponding to the line symmetry of the first region 1RE with the center line CL as the axis of symmetry, and a part of the symmetric region SR overlaps the first region 1RE. Thus, as shown in FIG. 4, by comparing the size of the overlapping region OLR where the symmetric region SR and the first region 1RE overlap with the determination threshold, the determination function 34c determines whether the first region 1RE is unilateral or bilateral.

[0035] Note that the processing circuit 34 determines, by the determination function 34c, whether the first region 1RE includes an anatomical site located beyond the center line CL with respect to the imaging site. When the first region 1RE includes an anatomical site, the processing circuit 34 may determine that the first region 1RE is bilateral, and when the first region 1RE does not include an anatomical site, the processing circuit 34 may determine that the first region 1RE is unilateral. When the imaging site is the brain, the anatomical site is, for example, the cerebellum, the brain stem, etc.

[0036] Specifically, prior to the determination by the determination function 34c, the processing circuit 34 specifies an anatomical site by performing predetermined image processing on the second medical image by means of the specific function 34b. As the predetermined image processing, for example, segmentation processing such as brain parcellation (brain differentiation), and known methods such as a learned model are applicable, and thus the description thereof is omitted. Note that brain parcellation corresponds to, for example, division of brain regions according to brain functions. At this time, the specific function 34b specifies a plurality of divided regions according to brain functions by performing brain parcellation on a brain image. Next, the determination function 34c compares the first region 1RE after application of the mapping matrix with the specified anatomical site, and determines that the first region 1RE is bilateral if the first region 1RE is included in the anatomical site, and determines that the first region 1RE is unilateral if the first region 1RE is not included in the anatomical site.

[0037] The processing circuit 34 determines a second region 2RE indicating a normal part at the imaging site by means of the determination function 34d. The second region 2RE is a region different from the first region 1RE in the second medical image 2ME and corresponds to a normal region indicating normal tissue. For example, when it is determined that the first region 1RE is unilateral, the determination function 34d determines, as the second region 2RE, a region corresponding to the line symmetry of the first region 1RE with the center line CL for the imaging site in the second medical image 2ME as the axis of symmetry. As shown in FIG. 2, when the first region 1RE is unilateral, the second region 2RE corresponds to the symmetric region SR. Further, the processing circuit 34 that realizes the determination function 34d corresponds to a determination unit.

[0038] Note that the determination function 34d may determine the entire other region excluding the first region 1RE in the second medical image 2ME as the second region 2RE. At this time, the determination function 34d determines the second region 2RE, for example, by subtracting the first region 1RE from the second medical image 2ME. Further, the determination function 34d may determine the second region 2RE by inputting the second medical image 2ME and the first region 1RE to a learned model capable of outputting the second region 2RE.

[0039] When it is determined that the first region 1RE is bilateral, the processing circuit 34 determines, by means of the determination function 34d, a region different from the first region 1RE in the second medical image as the second region 2RE. FIG. 5 is a diagram showing an example of the first region 1RE and the second region 2RE when the first region 1RE is bilateral. The position and size of the second region 2RE with respect to the imaging site are not limited to those in FIG. 5 and can be arbitrarily set.

[0040] For example, when it is determined that the first region 1RE is bilateral, the processing circuit 34 may determine, by means of the determination function 34d, a region having an apparent diffusion coefficient (ADC value) exceeding a predetermined value in the second medical image as the second region 2RE. The predetermined value is, for example, an ADC threshold value. Note that the predetermined value may be set to a value larger than the ADC threshold value. Specifically, the determination function 34d determines a region exceeding the predetermined value (hereinafter referred to as the excess region) in the ADC map. Next, the determination function 34d applies a mapping matrix to the excess region to map the excess region onto the second medical image 2ME. Subsequently, the determination function 34d determines, in the second medical image, a region different from the first region 1RE among the mapped excess regions as the second region 2RE.

[0041] Note that the processing circuit 34 may apply a mapping matrix to the ADC map by means of the determination function 34d to map the ADC map onto the second medical image 2ME, and then apply a predetermined value to the ADC map mapped onto the second medical image 2ME to determine the excess region. At this time, the determination function 34d determines, in the second medical image, a region different from the first region 1RE among the determined excess regions as the second region 2RE.

[0042] Also, when it is determined that the first region 1RE is bilateral, the processing circuit 34 may determine, by the determination function 34d, in the second medical image, as the second region 2RE, a region that does not include the cerebrospinal fluid of the subject and the region outside the brain (i.e., the region where the imaging site is not displayed) among the regions different from the first region 1RE. Specifically, prior to the determination of the second region 2RE by the determination function 34d, the specification function 34b specifies the region of the cerebrospinal fluid and the outside of the brain by performing predetermined image processing on the second medical image. Since various segmentation processes and known methods such as a learned model can be applied as the predetermined image processing, the description thereof is omitted. Next, the determination function 34d determines, in the second medical image, as the second region 2RE, a region that does not include the cerebrospinal fluid of the subject and the region outside the brain and is different from the first region 1RE.

[0043] In addition, when it is determined that the first region 1RE is bilateral, the processing circuit 34 may determine, by the determination function 34d, as the second region 2RE, an excess region that does not include the cerebrospinal fluid of the subject and the region outside the brain and exceeds a predetermined value in the ADC map after applying the mapping matrix among the regions different from the first region 1RE in the second medical image.

[0044] Also, when it is determined that the first region 1RE is bilateral, the processing circuit 34 may determine, by the determination function 34d, in the second medical image, as the second region 2RE, a region similar to the tissue properties in the first region 1RE. For example, prior to the determination of the second region 2RE by the determination function 34d, the specification function 34b specifies, by performing predetermined image processing on the second medical image, a region (hereinafter referred to as a similar region) similar to the first region 1RE in terms of tissue properties. Since various segmentation processes such as brain segmentation and known methods such as a learned model can be applied as the predetermined image processing, the description thereof is omitted. Next, the determination function 34d determines, in the second medical image, as the second region 2RE, a similar region among the regions different from the first region 1RE.

[0045] In addition, when it is determined that the first region 1RE is bilateral, the processing circuit 34 may determine, by means of the determination function 34d, only specific tissues such as white matter and gray matter in a region different from the first region 1RE in the second medical image 2ME as the second region. Further, for example, when the first region 1RE is a cerebellar infarction, the determination function 34d may determine, in the second medical image, a region that is not the cerebellum as the second region. Since known methods such as segmentation processing for brain differentiation of the second medical image and a learned model can be applied for specifying specific tissues such as white matter and gray matter and specifying regions other than the cerebellum, the description thereof is omitted. Further, the determination function 34d may, for example, perform segmentation of the brain region in the second medical image using an automatic extraction application for the brain region in the second medical image and select the second region.

[0046] The processing circuit 34 calculates, by means of the calculation function 34e, a first feature amount indicating a feature based on a plurality of pixel values in the first region 1RE and a second feature amount indicating a feature based on a plurality of pixel values in the second region 2RE in the second medical image 2ME. The first feature amount is a statistical value based on a plurality of pixel values included in the first region 1RE in the second medical image 2ME. Further, the second feature amount is a statistical value based on a plurality of pixel values included in the second region 2RE in the second medical image 2ME. The statistical value is, for example, the median of a plurality of pixel values. That is, the first feature amount is the median (hereinafter referred to as the first median) of a plurality of pixel values included in the first region 1RE in the second medical image 2ME. Further, the second feature amount is the median (hereinafter referred to as the second median) of a plurality of pixel values included in the second region 2RE in the second medical image 2ME. Note that the statistical value corresponding to the second feature amount may be the standard deviation (hereinafter referred to as the normal standard deviation) based on a plurality of pixel values in the second region 2RE. Note that the statistical value is not limited to the above, and known ones can be appropriately used.

[0047] The processing circuit 34 calculates, by means of the computing function 34e, an index regarding the mismatch between the first medical image and the second medical image (hereinafter referred to as the mismatch index) based on the first feature amount and the second feature amount. The mismatch index is, for example, the ratio between the first feature amount and the second feature amount. Specifically, the computing function 34e calculates the ratio of the first median value to the second median value as the mismatch index. Note that the computing function 34e may calculate, as a map of the mismatch index in the first region 1RE, the ratio of each of the plurality of pixel values in the first region 1RE to the second median value. Further, when the second median value is 0, the computing function 34e may calculate, as a map of the mismatch index in the first region 1RE, a value indicating how many times each of the plurality of pixel values in the first region 1RE is the normal standard deviation. At this time, the value (mismatch index) of each of the maps of the mismatch index corresponds to a numerical value corresponding to the Z-score.

[0048] The processing circuit 34 stores, in association with the first region 1RE, the mismatch index or the map of the mismatch index calculated by the computing function 34e in the memory 33. The processing circuit 34 that realizes the computing function 34e corresponds to the calculation unit.

[0049] The processing circuit 34 reads out and executes a program corresponding to the display control function 34f from the memory 33. As a result, the display control function 34f causes the display 32 to display the mismatch index calculated by the computing function 34e, the first region 1RE, and the second region 2RE. For example, the display control function 34f causes the display 32 to display a superimposed image in which the first region 1RE, the second region 2RE, and the mismatch index are superimposed on the second medical image. At this time, the display 32 superimposes the first region 1RE and the second region 2RE on the second medical image 2ME and displays them together with the mismatch index. The processing circuit 34 that realizes the display control function 34f corresponds to the display control unit.

[0050] The overall configuration of the medical information processing system 1 according to the embodiment has been described above. Hereinafter, processing related to the quantitative evaluation of mismatches in the first medical image and the second medical image (hereinafter referred to as mismatch evaluation processing) will be described with reference to FIG. 6.

[0051] FIG. 6 is a flowchart showing an example of the procedure of mismatch evaluation processing. Hereinafter, for the sake of specific description, the imaging site is the brain of the subject as described above, the first medical image is a diffusion-weighted image, and the second medical image is a FLAIR image. Also, the first medical image and the second medical image are assumed to have been generated in advance prior to the execution of the mismatch evaluation processing.

[0052] (Mismatch Evaluation Processing) (Step S601) The processing circuit 34 acquires the first medical image and the second medical image from the MRI apparatus 10 or the image storage apparatus 20 by the acquisition function 34a. The acquisition function 34a stores the first medical image and the second medical image in the memory 33.

[0053] (Step S602) The processing circuit 34 executes image processing on the first medical image by the specifying function 34b and specifies a first region in the first medical image. For example, the specifying function 34b specifies, as the first region, a plurality of pixels having pixel values lower than the ADC threshold value in the ADC map in the first medical image. The specifying function 34b stores the first region 1RE specified on the second medical image by the mapping matrix in the memory 33.

[0054] (Step S603) The processing circuit 34 determines, by the determination function 34c, whether the first region is bilateral or unilateral with respect to the imaging site based on the position of the first region in the imaging site. For example, if the first region straddles the axis of symmetry (center line) in the second medical image, the determination function 34c determines the first region as bilateral. Other determination methods will not be described since they conform to the above description.

[0055] (Step S604) If the first region is determined to be bilateral (Yes in Step S604), the process of Step S605 is executed. If the first region 1RE is not determined to be bilateral (No in Step S604), that is, if the first region is determined to be unilateral, the process of Step S606 is executed.

[0056] (Step S605) The processing circuit 34 determines, by the determination function 34d, a region different from the first region 1RE in the second medical image 2RE as the second region 2RE. Since the method for determining the second region 2RE when the first region is determined to be bilateral conforms to the above, the description is omitted. The determination function 34d stores the second region 2RE determined on the second medical image in the memory 33.

[0057] (Step S606) The processing circuit 34 determines, by the determination function 34d, a region (target region) corresponding to the line symmetry of the first region 1RE as the second region 2RE as the symmetry axis of the imaging site in the second medical image 2RE. Since the method for determining the second region 2RE when the first region 1RE is determined to be unilateral conforms to the above description, the description is omitted. The determination function 34d stores the second region 2RE determined on the second medical image in the memory 33.

[0058] (Step S607) The processing circuit 34 calculates, by the calculation function 34e, a first feature amount for the first region 1RE and a second feature amount for the second region 2RE in the second medical image 2RE. Since the processing content regarding the calculation of the first feature amount and the second feature amount conforms to the above content, the description is omitted. The calculation function 34e calculates a mismatch index based on the first feature amount and the second feature amount. Also, since the processing content regarding the calculation of the mismatch index conforms to the above content, the description is omitted. The calculation function 34e stores the mismatch index in the memory 33 in association with the first region 1RE and the second region 2RE.

[0059] (Step S608) The processing circuit 34, by means of the display control function 34f, superimposes the first region 1RE and the second region 2RE on the second medical image 2ME and causes the display 32 to display them together with the calculated mismatch index. Hereinafter, a display example when the first region 1RE is unilateral and a display example when the first region 1RE is bilateral will be described with reference to FIGS. 7 to 9.

[0060] FIG. 7 is a diagram showing an example of being displayed on the display 32 with the first region 1RE and the second region 2RE superimposed on the second medical image 2ME and together with the legend INB of the mismatch index when the first region 1RE is determined to be unilateral. In FIG. 7, the ratio of the first median value to the second median value is used as the mismatch index, superimposed on the first region 1RE, and displayed on the display 32.

[0061] FIG. 8 is a diagram showing an example of being displayed on the display 32 with the first region 1RE and the second region 2RE superimposed on the second medical image 2ME and together with the legend INB of the mismatch index when the first region 1RE is determined to be bilateral. In FIG. 8, the ratio of the first median value to the second median value is used as the mismatch index, superimposed on the first region 1RE, and displayed on the display 32.

[0062] FIG. 9 is a diagram showing an example in which a plurality of divided regions are further superimposed on the display example shown in FIG. 8. For example, in FIG. 9, a plurality of divided regions (A1 to A6) are respectively superimposed on the second medical image 2ME, and then the first region 1RE and the second region 2RE are superimposed. As shown in FIG. 9, by superimposing a plurality of divided regions corresponding to the functions of the brain, the correspondence between the functions of the brain in the lesion region and the normal region can be easily grasped. Therefore, according to the display example shown in FIG. 9, the functional regions of the brain regarding the lesion region and the normal region can be easily grasped, so that the burden of the examination on the user can be reduced. From these facts, according to the medical image processing apparatus 30 according to the embodiment, the throughput of the examination for the subject can be improved.

[0063] The medical image processing apparatus 30 according to the above-described embodiment acquires a first medical image and a second medical image 2ME, identifies a first region 1RE based on the first medical image, and determines whether the first region 1RE is bilateral or unilateral based on the position of the first region 1RE in the imaging region in the second medical image 2RE. When it is determined that the first region 1RE is bilateral, a region different from the first region 1RE in the second medical image 2RE is determined as a second region 2RE. When it is determined that the first region 1RE is unilateral, a region corresponding to the line symmetry of the first region 1RE with the center line CL of the imaging region as the axis of symmetry in the second medical image 2ME is determined as the second region 2RE. A first feature amount related to the first region 1RE and a second feature amount related to the second region 2RE are calculated in the second medical image, and a mismatch index is calculated based on the first feature amount and the second feature amount. Further, the medical image processing apparatus 30 according to the embodiment superimposes the first region 1RE and the second region 2RE on the second medical image 2ME and displays them together with the mismatch index.

[0064] In the medical image processing apparatus 30 according to the embodiment, the first medical image and the second medical image are an axial cross-section of the imaging region or a coronal cross-section of the imaging region. Further, in the medical image processing apparatus 30 according to the embodiment, the first medical image is a diffusion-weighted image, and the second medical image is a FLAIR image. Further, in the medical image processing apparatus 30 according to the embodiment, the first feature amount is a statistical value based on a plurality of pixel values included in the first region 1RE, and the second feature amount is a statistical value based on a plurality of pixel values included in the second region 2RE. Further, in the medical image processing apparatus 30 according to the embodiment, the mismatch index is a ratio between the first feature amount and the second feature amount.

[0065] When the symmetric region SR corresponding to the line symmetry of the first region 1RE and the first region 1RE overlap with the central axis of the imaging site as the axis of symmetry, the medical image processing apparatus 30 according to the embodiment may determine that the first region 1RE is bilateral. When the symmetric region SR and the first region 1RE do not overlap, the medical image processing apparatus 30 according to the embodiment may determine that the first region 1RE is unilateral. Further, when the size of the overlapping region OLR where the symmetric region SR and the first region 1RE overlap exceeds a predetermined threshold, the medical image processing apparatus 30 according to the embodiment may determine that the first region 1RE is bilateral. When the size of the overlapping region OLR is equal to or less than the predetermined threshold, the medical image processing apparatus 30 according to the embodiment may determine that the first region 1RE is unilateral.

[0066] Further, the medical image processing apparatus 30 according to the embodiment identifies an anatomical site located beyond the central line CL of the imaging site by performing predetermined image processing on the second medical image 2ME, and determines whether or not the first region 1RE is included in the identified anatomical site. When the first region 1RE is included in the anatomical site, the medical image processing apparatus 30 according to the embodiment may determine that the first region 1RE is bilateral. When the first region 1RE is not included in the anatomical site, the medical image processing apparatus 30 according to the embodiment may determine that the first region 1RE is unilateral.

[0067] Further, when it is determined that the first region 1RE is bilateral, the medical image processing apparatus 30 according to the embodiment may determine a region having an apparent diffusion coefficient exceeding a predetermined value as the second region 2RE. Further, in the medical image processing apparatus 30 according to the embodiment, the imaging site is the brain of the subject. When it is determined that the first region 1RE is bilateral, the medical image processing apparatus 30 according to the embodiment may determine a region that does not include the cerebrospinal fluid of the subject and the region outside the brain as the second region 2RE. Further, when it is determined that the first region 1RE is bilateral, the medical image processing apparatus 30 according to the embodiment may determine a region similar to the properties of the tissue in the first region 1RE as the second region 2RE.

[0068] From these, according to the medical image processing apparatus 30 according to the embodiment, it is possible to determine whether a lesion (for example, cerebral infarction) of a subject is unilateral or bilateral, and appropriately set a healthy-side region (normal region) that serves as a criterion for calculating a mismatch index for determining whether the lesion region is in the acute phase. Thereby, according to the medical image processing apparatus 30 according to the embodiment, regardless of the bilaterality and unilaterality of the lesion, the accuracy of the quantitative evaluation of the DWI / FLAIR mismatch phenomenon can be improved. In addition, according to the medical image processing apparatus 30 according to the embodiment, it is not necessary to have a doctor such as a user qualitatively judge the DWI / FLAIR mismatch phenomenon in his or her own mind, and the burden on the user for the examination can be reduced. Furthermore, according to the medical image processing apparatus 30 according to the embodiment, since the qualitative judgment by the user can be reduced, the throughput of the examination for the subject can be improved.

[0069] (Application Example) This application example is to determine whether a lesion is in the acute phase based on a mismatch index and display the determination result on the display 32. The memory 33 stores a threshold value (hereinafter referred to as an acute-phase determination threshold value) for determining whether the mismatch index is in the acute phase. In the mismatch evaluation process in this application example, the following process is executed following step S607 in FIG. 6. The processing circuit 34 compares the mismatch index with the acute-phase determination threshold value by the determination function 34c. When the mismatch index exceeds the acute-phase determination threshold value, the determination function 34c determines that the lesion related to the first region 1RE is in the acute phase.

[0070] Also, in the mismatch evaluation process in the application example, in step S608 in FIG. 6, the following process is further executed. The processing circuit 34 causes the display control function 34f to display the mismatch index and the determination result of whether it is in the acute phase on the display 32 together with a superimposed image in which the first region 1RE and the second region 2RE are superimposed on the second medical image 2ME.

[0071] Incidentally, as a modification of this application example, the memory 33 may store an algorithm such as a calculation formula or a learned model (hereinafter referred to as a lesion onset calculation algorithm) that calculates the onset time of a lesion related to the first region 1RE using the mismatch index as an input and / or the elapsed time from the onset of the lesion. At this time, the processing circuit 34 calculates the onset time of the lesion and / or the elapsed time from the onset of the lesion by applying the mismatch index to the lesion onset calculation algorithm by means of the computing function 34e. Next, the processing circuit 34 causes the display control function 34f to display on the display 32 the onset time of the lesion and / or the elapsed time from the onset of the lesion, in addition to the superimposed image in which the first region 1RE and the second region 2RE are superimposed on the second medical image 2ME, the mismatch index, and the determination result as to whether it is the acute phase or not.

[0072] Incidentally, instead of the lesion onset calculation algorithm, a correspondence table of the onset time of the lesion with respect to the mismatch index and / or the elapsed time from the onset of the lesion may be stored in the memory 33. At this time, the processing circuit 34 determines the onset time of the lesion and / or the elapsed time from the onset of the lesion by collating the mismatch index with the correspondence table by means of the determination function 34d.

[0073] According to the medical image processing apparatus 30 according to this application example, it is possible to predict the onset time and the like in unilateral and bilateral lesions (for example, cerebral infarction). Thereby, according to the medical image processing apparatus 30 according to this application example, the burden of the examination on the user can be further reduced, and the throughput of the examination on the subject can be further improved. Since other effects are the same as those of the embodiment, the description thereof is omitted.

[0074] When realizing the technical idea in this embodiment by a medical image processing method, the medical image processing method acquires a first medical image collected by a predetermined imaging of an imaging part of a subject, and a second medical image 2RE collected by an imaging different from the predetermined imaging and including the imaging part. Based on the first medical image, a first region 1RE related to a lesion in the imaging part is specified. Based on the position of the first region 1RE in the imaging part, it is determined whether the first region 1RE is bilateral or unilateral with respect to the imaging part. When it is determined that the first region 1RE is bilateral, a region different from the first region 1RE in the second medical image 2RE is determined as a second region 2RE indicating a normal part in the imaging part. When it is determined that the first region 1RE is unilateral, a region corresponding to the line symmetry of the first region 1RE with the center line CL of the imaging part in the second medical image 2ME as the axis of symmetry is determined as the second region 2RE. In the second medical image 2ME, a first feature amount indicating a feature by a plurality of pixel values in the first region 1RE and a second feature amount indicating a feature by a plurality of pixel values in the second region 2RE are calculated. Based on the first feature amount and the second feature amount, an index related to the mismatch between the first medical image and the second medical image 2ME is calculated. The processing procedure in the medical image processing method conforms to the procedure of the mismatch evaluation processing. Also, the effect by the medical image processing method is the same as that of the embodiment. From these, the description of the processing procedure and the effect of the mismatch evaluation processing in the medical image processing method is omitted.

[0075] When the technical idea in the embodiment is realized by a medical image processing program, the medical image processing program causes a computer to acquire a first medical image collected by a predetermined imaging of an imaging site of a subject and a second medical image 2RE collected by an imaging different from the predetermined imaging and including the imaging site, identify a first region 1RE related to a lesion in the imaging site based on the first medical image, determine whether the first region 1RE is bilateral or unilateral with respect to the imaging site based on the position of the first region 1RE in the imaging site, when it is determined that the first region 1RE is bilateral, determine a region different from the first region 1RE in the second medical image 2RE as a second region 2RE indicating a normal part in the imaging site, when it is determined that the first region 1RE is unilateral, determine a region corresponding to the line symmetry of the first region 1RE with the center line CL with respect to the imaging site in the second medical image 2ME as the second region 2RE, calculate a first feature amount indicating a feature by a plurality of pixel values in the first region 1RE and a second feature amount indicating a feature by a plurality of pixel values in the second region 2RE in the second medical image 2ME, and calculate an index related to the mismatch between the first medical image and the second medical image 2ME based on the first feature amount and the second feature amount.

[0076] For example, the mismatch evaluation process can also be realized by installing an image processing program on a computer such as the medical image processing device 30 or the MRI device 10 shown in FIG. 1 and expanding them in the memory. At this time, the program that can cause the computer to execute the process can also be stored and distributed in a storage medium such as a magnetic disk (such as a hard disk), an optical disk (such as a CD-ROM, DVD), or a semiconductor memory. Further, the distribution of the medical image processing program is not limited to the above media, and may be distributed using a telecommunication function such as downloading via the Internet. The processing procedure in the medical image processing program conforms to the mismatch evaluation process. Also, the effects of the medical image processing program are the same as those in the embodiment. For these reasons, the description of the processing procedure and effects of the mismatch evaluation process in the medical image processing program is omitted.

[0077] The technical features in this embodiment can be realized by an MRI device. At this time, the processing circuit mounted on the MRI device has the acquisition function 34a, the specification function 34b, the determination function 34c, the decision function 34d, the calculation function 34e, and the display control function 34f shown in FIG. 1. At this time, the MRI device realizes the mismatch evaluation process. The processing procedure in the MRI device that realizes the acquisition function 34a, the specification function 34b, the determination function 34c, the decision function 34d, the calculation function 34e, and the display control function 34f conforms to the mismatch evaluation process of the embodiment. Also, the effects of the MRI device are the same as those of the embodiment. For these reasons, the description of the processing procedure and effects of the mismatch evaluation process in the MRI device is omitted.

[0078] According to at least the embodiments, application examples, etc. described above, regardless of the bilateral or unilateral nature of the lesion, it is possible to improve the accuracy of the quantitative evaluation of the mismatch in two medical images.

[0079] Although several embodiments have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, changes, and combinations of the embodiments can be made without departing from the gist of the invention. These embodiments and their modifications are included in the scope and gist of the invention, and are also included in the invention described in the claims and the equivalent scope thereof.

Explanation of Reference Numerals

[0080] 1 Medical information processing system 10 MRI device 20 Image storage device 30 Medical image processing device 31 Input interface 32 Display 33 Memory 34 Processing circuit 34a Acquisition function 34b Specification function 34c determination function 34d decision function 34e calculation function 34f display control function

Claims

1. An acquisition unit that acquires a first medical image collected by a predetermined imaging of an imaging site of a subject and a second medical image collected by an imaging different from the predetermined imaging and including the imaging site; A specifying unit that specifies a first region related to a lesion in the imaging site based on the first medical image; A determination unit that determines whether the first region is bilateral or unilateral with respect to the imaging site based on the position of the first region in the imaging site; When it is determined that the first region is bilateral, a region different from the first region in the second medical image is determined as a second region indicating a normal part in the imaging site, and when it is determined that the first region is unilateral, a region corresponding to the line symmetry of the first region with the center line of the imaging site as the axis of symmetry in the second medical image is determined as the second region; A calculation unit that calculates a first feature amount indicating a feature by a plurality of pixel values in the first region and a second feature amount indicating a feature by a plurality of pixel values in the second region in the second medical image, and calculates an index related to the mismatch between the first medical image and the second medical image based on the first feature amount and the second feature amount; A medical image processing apparatus comprising the above.

2. The first medical image and the second medical image are an axial cross-section of the imaging site or a coronal cross-section of the imaging site. The medical image processing apparatus according to Claim 1.

3. The first medical image is a diffusion-weighted image, The second medical image is a FLAIR (Fluid Attenuated Inversion Recovery) image. The medical image processing apparatus according to Claim 1.

4. The first feature amount is a statistical value based on a plurality of pixel values included in the first region, The second feature amount is a statistical value based on a plurality of pixel values included in the second region. The medical image processing apparatus according to Claim 1.

5. The index is a ratio between the first feature amount and the second feature amount. The medical image processing apparatus according to Claim 1.

6. When the symmetric region corresponding to the line symmetry of the first region with the center line of the imaging site as the axis of symmetry overlaps with the first region, the determination unit determines that the first region is bilateral, and when the symmetric region and the first region do not overlap, the determination unit determines that the first region is unilateral. The medical image processing apparatus according to any one of claims 1 to 5.

7. When the size of the overlapping region where the symmetric region corresponding to the line symmetry of the first region overlaps with the first region with the center line with respect to the imaging site as the axis of symmetry exceeds a predetermined threshold value, the determination unit determines that the first region is bilateral, and when the size of the overlapping region is equal to or less than the predetermined threshold value, the determination unit determines that the first region is unilateral. The medical image processing apparatus according to any one of claims 1 to 5.

8. The determination unit determines whether or not the first region is included in an anatomical site located beyond the center line with respect to the imaging site. When the first region is included in the anatomical site, the determination unit determines that the first region is bilateral, and when the first region is not included in the anatomical site, the determination unit determines that the first region is unilateral. The medical image processing apparatus according to any one of claims 1 to 5.

9. The specifying unit specifies the anatomical site by performing predetermined image processing on the second medical image. The medical image processing apparatus according to claim 8.

10. When it is determined that the first region is bilateral, the determination unit determines a region having an apparent diffusion coefficient exceeding a predetermined value as the second region. The medical image processing apparatus according to any one of claims 1 to 5.

11. The imaging site is the brain of the subject, When it is determined that the first region is bilateral, the determination unit determines a region that does not include the cerebrospinal fluid of the subject and the region outside the brain as the second region. The medical image processing apparatus according to any one of claims 1 to 5.

12. When it is determined that the first region is bilateral, the determination unit determines a region similar to the property of the tissue in the first region as the second region. The medical image processing apparatus according to any one of claims 1 to 5.

13. The apparatus further includes a display unit that superimposes the first region and the second region on the second medical image and displays them together with the index. The medical image processing apparatus according to any one of claims 1 to 5.

14. A first medical image collected by predetermined imaging of an imaging site of a subject and a second medical image collected by imaging different from the predetermined imaging and including the imaging site are acquired. Based on the first medical image, a first region related to a lesion in the imaging site is specified. Based on the position of the first region in the imaging region, determine whether the first region is bilateral or unilateral with respect to the imaging region. When it is determined that the first region is bilateral, determine a region different from the first region in the second medical image as a second region indicating a normal part in the imaging region. When it is determined that the first region is unilateral, determine a region corresponding to the line symmetry of the first region with the center line with respect to the imaging region in the second medical image as the second region. In the second medical image, calculate a first feature amount indicating a feature based on a plurality of pixel values in the first region and a second feature amount indicating a feature based on a plurality of pixel values in the second region. Calculate an index related to the mismatch between the first medical image and the second medical image based on the first feature amount and the second feature amount. A medical image processing method comprising the above.

15. On a computer, Obtain a first medical image collected by a predetermined imaging of an imaging region of a subject and a second medical image collected by an imaging different from the predetermined imaging and including the imaging region. Based on the first medical image, identify a first region related to a lesion in the imaging region. Based on the position of the first region in the imaging region, determine whether the first region is bilateral or unilateral with respect to the imaging region. When it is determined that the first region is bilateral, determine a region different from the first region in the second medical image as a second region indicating a normal part in the imaging region. When it is determined that the first region is unilateral, determine a region corresponding to the line symmetry of the first region with the center line with respect to the imaging region in the second medical image as the second region. In the second medical image, calculate a first feature amount indicating a feature based on a plurality of pixel values in the first region and a second feature amount indicating a feature based on a plurality of pixel values in the second region. Calculate an index related to the mismatch between the first medical image and the second medical image based on the first feature amount and the second feature amount. A medical image processing program for realizing the above.

Images