JP2020062215A - Medical image browsing device, medical image processing device and medical image diagnostic device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent an individual from being identified and identified from a medical image without impairing the original purpose of the medical image such as diagnosis and treatment. A medical image viewing device according to an embodiment includes a display, a generation unit, a determination unit, and a processing unit. The generation unit generates a display image to be displayed on the display from the medical image data in which the subject is imaged according to the user's instruction. The determination unit determines whether or not the individual of the subject can be identified from the display image each time the display state of the display image is changed according to the user instruction. When it is determined that an individual can be identified from the displayed image, the processing unit performs anonymization processing for preventing the individual from being identified from the displayed image in real time each time the display state of the displayed image is changed. , The display image after the anonymization process is displayed on the display. [Selection diagram] Fig. 2

Description

  Embodiments of the present invention relate to a medical image browsing apparatus, a medical image processing apparatus, and a medical image diagnostic apparatus.

  Image data generated by imaging a subject with a modality such as an X-ray CT (Computed Tomography) apparatus, an MRI (Magnetic Resonance Imaging) apparatus, or an ultrasonic diagnostic apparatus is, for example, PACS (Picture Archiving and Communication Systems). It is sent to a so-called medical image management system and stored in the image server in PACS.

  The image data stored in the image server in the PACS is read by a doctor in charge of imaging or a person involved in hospital such as an image interpretation doctor via a medical image browsing device (so-called image viewer). The medical image browsing apparatus performs image processing on the read image data based on an instruction from a user (a doctor in charge or an image interpreting doctor). The image processing here is, for example, processing of rotating three-dimensional image data in a desired direction to generate a volume rendering image, generation of a MIP (Maximum Intensity Projection) image, or three-dimensional image. It includes a process of cutting out data in a desired slice section (so-called MPR method: Multi Planer Reconstruction) and the like. The images generated by these image processes are displayed on the display provided in the medical image browsing apparatus, and are used for diagnosis by a medical doctor in charge and interpretation by an image interpretation doctor and used for diagnosis and treatment of patients.

  On the other hand, image data acquired by various modalities are used for diagnosis and treatment of patients, research in hospitals, research in various medical research institutes including university medical departments and hospitals, and medical-related It is often used for academic conference presentations. It may also be used as a teaching material for medical students.

  As described above, the image data of the subject (for example, patient) imaged by the modality potentially has a possibility of being provided to a large number of observers.

  On the other hand, there is a remarkable progress in high-definition technology for modalities such as X-ray CT apparatuses and MRI apparatuses today. As a result, for example, it has become possible to easily specify an individual to be imaged from a three-dimensional image of the head acquired by the X-ray CT apparatus. For this reason, the technology of concealing personal information such as the patient's name and address, that is, the personal information as text information is not sufficient, and there is a demand for technology that makes it impossible to identify an individual from the medical image itself. There is.

  For example, by detecting the body surface of the subject from the three-dimensional image data and processing the pixel value of the portion of the image data having a predetermined thickness from the body surface by masking or mosaic processing, that is, the three-dimensional image. A technique has been proposed in which it is difficult to discriminate an individual from the generated three-dimensional image even when the three-dimensional image is generated by three-dimensional image processing such as volume rendering by processing the data itself in advance.

  However, the original purpose of medical image data is to diagnose a subject and support treatment. Therefore, if the pixel value of a predetermined thickness part from the surface of the face is processed by mask processing or mosaic processing in order to make it difficult to identify an individual, face parts such as eyes, nose, ears, mouth, and the head, and the head. The diagnosis and treatment of the department will be seriously hindered.

JP, 2013-138800, A

  The problem to be solved by the present invention is to prevent an individual from being discriminated and specified from a medical image without impairing the original purpose of the medical image such as diagnosis and treatment.

  The medical image browsing device according to the embodiment includes a display, a generation unit, a determination unit, and a processing unit. The generation unit generates a display image to be displayed on the display from medical image data obtained by imaging the subject according to a user instruction. The determination unit determines whether or not the individual of the subject can be identified from the display image every time the display state of the display image is changed in response to the user instruction. When it is determined that the individual can be identified from the display image, the processing unit performs anonymization processing that prevents the individual from being identified from the display image in real time every time the display state of the display image is changed. The display image after the anonymization process is displayed on the display.

FIG. 1 is a diagram showing a configuration example of a medical image processing system including a medical image browsing device according to a first embodiment. The block diagram which shows the structural example of the medical image browsing apparatus which concerns on 1st Embodiment. 3 is a flowchart showing an operation example of the medical image browsing apparatus according to the first embodiment. The figure explaining the concept of the process until image data is acquired and a display image is produced | generated. The figure which illustrates the determination table for demonstrating the concept of determination processing. The figure explaining the 1st example of determination processing and anonymization processing. The figure explaining the 2nd example of determination processing and anonymization processing. The figure explaining the 3rd example of determination processing and anonymization processing. The figure explaining the 4th example of determination processing and anonymization processing. The figure explaining the 5th example of determination processing and anonymization processing. The figure explaining the 6th example of determination processing and anonymization processing. The flowchart which shows the operation example of the medical image browsing apparatus of 2nd Embodiment. The flowchart which shows the operation example of the medical image browsing apparatus of 3rd Embodiment. The block diagram which shows the structural example of the medical image processing apparatus which concerns on other embodiment. The block diagram which shows the structural example of the medical image diagnostic apparatus which concerns on other embodiment.

  A medical image browsing apparatus, a medical image processing apparatus, and a medical image diagnostic apparatus according to embodiments will be described with reference to the accompanying drawings. FIG. 1 is a diagram showing a configuration example of a medical image processing system including a medical image browsing apparatus 100, a medical image diagnostic apparatus 510, and a medical image processing apparatus 530 according to an embodiment. The medical image processing system is, for example, a series of processing systems related to medical images that acquire, image-process, store, and browse medical images in a hospital.

  The medical image diagnostic apparatus 510 is an apparatus that images a subject such as a patient and performs processing such as reconstruction on the data collected by the imaging to generate a medical image. The medical image diagnostic apparatus 510 is also called a modality 510 and includes an X-ray CT apparatus 511, an MRI apparatus 512, an ultrasonic diagnostic apparatus 513, and the like.

  The image server 520 is a device that stores the medical images and image data acquired and generated by the medical image diagnostic apparatus 510. The medical image processing apparatus 530 performs image processing on the medical image or image data read from the image server 520 or the medical image or image data acquired from the medical image diagnostic apparatus 510, and the processed medical image. Is stored in the image server 520.

  The medical image browsing apparatus 100 performs image processing on the medical image or image data read from the image server 520 or the medical image or image data acquired from the medical image diagnostic apparatus 510 to obtain the medical image browsing apparatus. It is a device that generates a display image to be displayed on a display device included in 100 and causes the display device to display it.

  The medical image diagnostic apparatus 510, the image server 500, the medical image processing apparatus 530, and the medical image browsing apparatus 100 are connected to each other, for example, via a network 500 in a hospital so that various data and medical images can be exchanged. There is.

  Of the medical image browsing apparatus 100, the medical image processing apparatus 530, and the medical image diagnostic apparatus 510 according to the embodiment, first, the medical image browsing apparatus 100 will be taken up and the configuration and operation thereof will be described below. After that, the medical image processing apparatus 530 and the medical image diagnostic apparatus 510 will be briefly described.

(Configuration of the first embodiment)
FIG. 2 is a block diagram showing a configuration example of the medical image browsing apparatus 100 according to the first embodiment. The medical image browsing apparatus 100 includes, for example, a communication interface circuit 10, a recording medium interface circuit 11, a processing circuit 20, a storage circuit 30, an input device 40, and a display 50. The medical image browsing apparatus 100 is configured as, for example, a so-called workstation or a high performance personal computer.

  The communication interface circuit 10 is an interface circuit that inputs / outputs data via a wired or wireless local network, a wide area network, the Internet, or a dedicated or general-purpose communication line. The recording medium interface circuit 11 is an interface circuit that inputs / outputs data via a storage medium such as an optical disk, a USB memory, an SD memory card.

  In the medical image browsing apparatus 100 according to the first embodiment, medical image data of an image of a subject is input via the communication interface circuit 10 or the recording medium interface circuit 11.

  The storage circuit 30 is a storage medium including a ROM (Read Only Memory) and a RAM (Random Access Memory) as well as an external storage device such as a HDD (Hard Disk Drive) and an optical disk device. The memory circuit 30 stores data necessary for realizing various functions described below, and also stores various programs executed by a processor included in the processing circuit 20.

  The input device 40 is, for example, a mouse, a keyboard, a trackball, a touch panel, or the like, and includes various devices for the operator to input various information and data.

  The display 50 is a display device such as a liquid crystal display panel, a plasma display panel, and an organic EL panel.

  The processing circuit 20 is, for example, a circuit including a CPU and a dedicated or general-purpose processor. The processor implements various functions described below by executing various programs stored in the storage circuit 30. The processing circuit 20 may be configured by hardware such as FPGA (field programmable gate array) or ASIC (application specific integrated circuit). Various functions described below can also be realized by these hardware. Further, the processing circuit 20 can realize various functions by combining software processing by a processor and a program, and hardware processing.

  Alternatively, the processing circuit 20 may be configured by combining a plurality of independent processors, and each processor may realize each function. When a plurality of processors are provided, the storage medium for storing the program may be provided individually for each processor, or one storage medium may collectively store the programs corresponding to the functions of all the processors. Good.

(Operation of the first embodiment)
The processing circuit 20 of the first embodiment realizes the respective functions shown in FIG. 2, that is, the image data acquisition function 21, the image processing function 22, the determination function 23, and the anonymization processing function 24. Each of these functions is realized, for example, by a processor included in the processing circuit 20 executing a predetermined program stored in the storage circuit 30.

  The image data acquisition function 21 is a function of taking in the image data input via the communication interface circuit 10 or the recording medium interface circuit 11 to the processing circuit 20. The image data acquired by the image data acquisition function 21 is, for example, medical image data in which a region including the head of the subject is imaged. The type of image data is, for example, three-dimensional image data captured by the X-ray CT apparatus 511, the MRI apparatus 512, or the ultrasonic diagnostic apparatus 513, but is not limited to these, and two-dimensional X-ray image data. But it's okay.

  The image processing function 22 generates a display image to be displayed on the display 50 from the input image data according to a user instruction. For example, when the MPR image is generated as the display image, the MPR image is generated from the three-dimensional image data according to the position and the direction of the cutting section based on the user's instruction.

  Further, when a volume rendering image is generated as a display image, volume rendering processing is performed on the three-dimensional image data according to the projection direction based on the user instruction, and the volume rendering image is generated. In this case, the projection direction is changed according to the rotation instruction from the user, and the volume rendering image is sequentially updated by following the change in the projection direction in almost real time. In addition, the enlargement, reduction, or movement processing for the volume rendering image is performed in substantially real time in response to a user's instruction for enlargement, reduction, or movement.

  The determination function 23 determines whether the individual of the subject can be identified from the display image. Then, every time the display state of the display image is changed according to the user instruction, it is determined whether or not the individual of the subject can be specified from the display image. More specific processing of the determination function 23 will be described later.

  The anonymization processing function 24 performs anonymization processing that prevents an individual from being identified from the display image when the determination function 23 determines that the individual can be identified from the display image. The determination of whether or not the individual of the subject can be identified from the display image is performed in real time every time the display state of the display image is changed, and accordingly, the anonymization process also determines the display state of the display image. Every time it is changed, it is done in real time. Further, the anonymization processing function 24 causes the display 50 to display the display image after the anonymization processing.

  FIG. 3 is a flowchart showing an operation example of the medical image browsing apparatus 100 according to the first embodiment. Further, FIG. 4 to FIG. 11 are diagrams illustrating the operation concept of the medical image browsing apparatus 100 according to the first embodiment. Hereinafter, the operation of the medical image browsing apparatus 100 will be described more specifically along the flow of FIG. 3 and with reference to FIGS. 4 to 11.

  First, in step ST100 of FIG. 3, the image data acquisition function 21 acquires the image data input via the communication interface circuit 10 or the recording medium interface circuit 11. Next, in step ST101, information regarding the user instruction is acquired. Further, in step ST102, a display image based on the user instruction is generated.

  FIG. 4 is a diagram illustrating the processing concept of the processing from the acquisition of image data to the generation of a display image (the processing from step ST100 to step ST102 described above). In step ST100, for example, image data as shown on the left side of FIG. 4 is acquired. The type of image data is not particularly limited, but in the following, for example, three-dimensional image data obtained by imaging the subject with the X-ray CT apparatus 511 will be described as an example of the image data acquired in step ST100. In the example shown in FIG. 4, three-dimensional image data of the upper half of the body of a man including the head is acquired.

  In step ST101 and step ST102, the image processing function 22 generates a display image according to a user instruction. For example, a user's instruction to enlarge the head of the three-dimensional image data of the upper half of the body of the man shown on the left side of FIG. The user instruction is given, for example, by operating a mouse or a touch panel by a user such as a doctor. By such a user instruction, the volume rendering image of the head of the man in the state of facing the front shown in the upper right part of FIG. 4 can be generated as a display image.

  Further, for example, by enlarging the head of the three-dimensional image data of the upper half of the body of the male shown on the left side of FIG. 4 and giving a user instruction to rotate the face so as to face the left side, the male shown in the right middle stage of FIG. It is possible to generate a volume rendering image of the head in a state in which is facing right to the left as a display image.

  Further, for example, volume rendering of the upper body from the neck shown in the lower right part of FIG. 4 is performed by instructing the user to cut out the upper body from the neck of the three-dimensional image data of the upper body of the male shown on the left side of FIG. The image can be generated as a display image.

  In step ST101 and step ST102, the user views the display image displayed on the display 50 and operates the mouse, the touch panel, or the like to enlarge, reduce, rotate, move, or cut the three-dimensional image data. An instruction is given, and a display image such as a volume rendering image is generated in real time in response to this user instruction.

  Returning to FIG. 3, in step ST103, it is determined whether or not an individual can be identified from the display image each time the display state of the display image is changed in accordance with a user instruction, not only the initial display state of the display image. .

  For example, the anatomical region necessary for identifying the individual is detected from the display image, and the determination index value for each region representing the ease of identifying the individual is calculated and calculated for each detected anatomical region. It is determined whether or not the individual can be identified from the display image using the region-specific determination index value.

  More specifically, the above-mentioned region-specific determination index value is calculated by weighting the basic value assigned to each anatomical region. Furthermore, the site-specific determination index values for each anatomical site calculated by weighting are added to obtain a total determination index value. Then, threshold determination is performed on this comprehensive determination index value to determine whether an individual can be identified from the display image.

  FIG. 5 is a determination table for explaining the concept of the determination process performed in step ST103. A part of this determination table may be stored in the storage circuit 30 and used for the determination process of step ST103.

  The first row of the determination table exemplifies “eyes”, “mouths”, “nose”, and “ears” as anatomical parts necessary for identifying an individual from the display image. These four anatomical parts are considered to be typical parts necessary for identifying an individual from the displayed image.

  The second row of the determination table shows the basic value assigned to each anatomical part. Here, the basic value is an index value indicating whether or not the corresponding anatomical part is visible in the display image. In the example of the determination table shown in FIG. 5, “25” points are uniformly assigned to “eye”, “mouth”, “nose”, and “ear”.

  The basic value does not indicate “how much the anatomical part in the display image is visible”, but if the corresponding anatomical part is slightly visible in the display image, a full point, for example, "25" points are assigned, and "0" points are assigned when the corresponding anatomical part is not visible at all in the display image.

  In other words, the basic value is an index value indicating whether or not the anatomical part is visible in the display image. Therefore, for example, in the display image shown on the right side of FIG. 4, in the example of the upper row in which the man faces the front, all of the “eyes”, “mouths”, “nose”, and “ears” are visible. All the four anatomical parts will be uniformly assigned with “25” points. In the display image shown on the right side of FIG. 4, also in the example of the middle stage in which the man is looking sideways, the anatomical parts of the “eye”, “mouth”, “nose”, and “ear” are displayed. Since at least a part is visible, “25” points are uniformly assigned to all four anatomical parts.

  On the other hand, of the display images shown on the right side of FIG. 4, in the lower example, the head itself is not visible, and of course, in this display image, “eyes”, “mouths”, “nose”, And neither anatomy of the "ear" is visible. Therefore, in the lower display image on the right side of FIG. 4, “0” is uniformly applied to all four anatomical parts of “eye”, “mouth”, “nose”, and “ear”. Points will be allocated.

  To determine whether or not anatomical parts such as “eyes”, “mouths”, “nose”, and “ears” are visible in the display image, for example, a known method such as template matching or machine learning is used. can do. In template matching, it is possible to determine whether or not each anatomical part is present in the display image by performing template matching with the display image and the template corresponding to each shape of the anatomical part. . In machine learning, the display image is applied in advance to the learned model learned by the teacher image including the “eye”, “mouth”, “nose”, “ear”, etc. It can be determined whether or not the target portion is present in the display image.

  The third row and the fourth row of the determination table show the area ratio weight W1 and the transparency weight W2 as weights indicating the exposure degree of the anatomical part in the display image. The weight W indicating the degree of exposure is the product of the area ratio weight W1 and the transparency weight W2.

  The area ratio weight W1 is the ratio of the area of the anatomical site that is visible in the display state of the display image to the area of the anatomical site when the entire anatomical site is visible. For example, when the area of both eyes facing the front is A in a display image with a certain magnification, and when only one eye facing the front is visible in the display image with the same magnification, the display of the display image is displayed. The area of the eyes in the state is A / 2. Therefore, in this case, the area ratio weight W1 is 0.5.

  The transparency weight W2 is a weight corresponding to the transparency value set for the anatomical parts such as “eyes”, “mouths”, “nose”, and “ears” in the display image. That is, when the corresponding anatomical part is completely transparent and the tissue therebelow is completely transparent and visible, the transparency weight W2 is 0.0. On the contrary, when the corresponding anatomical part is completely opaque and the tissue thereunder is completely invisible, the transparency weight W2 is 1.0.

  The area ratio weight W1 and the transparency weight W2 are determined by the determination function 23 of the processing circuit 20 based on the display image and its display state.

  The fifth row of the determination table shows the individual weight W3. The individuality weight W3 is a weight representing the strength of individuality indicating how different the shape of the corresponding anatomical part is from the standard shape. The individuality weight W3 is set by the user, for example.

  The sixth row of the determination table shows the determination index value for each part. The region-specific determination index value is a value obtained by weighting each basic value of the anatomical region with the area ratio weight W1, the transparency weight W2, and the individuality weight W3, that is, the basic value multiplied by each weight. .

  The seventh row of the determination table shows the total determination index value obtained by adding the determination index values for each part of the anatomical part.

  By performing threshold value determination using a predetermined threshold value on the comprehensive determination index value calculated in this way, it is determined whether or not to perform anonymization processing on the display image in the display state at that time. . That is, when the comprehensive determination index value exceeds the threshold value by the above determination in step ST103, it is determined that the individual can be identified from the display image in the display state, and the process proceeds to step ST104 to display the display image in the display state. Anonymization processing is performed on. Thereafter, in step ST105, the anonymized display image is displayed on the display 50. The processing of step ST104 and step ST105 is performed by the anonymization processing function 24 of FIG.

  On the other hand, if the comprehensive determination index value does not exceed the threshold value by the determination in step ST103, it is determined that the individual cannot be identified from the display image in the display state, and the anonymization process is not performed in step ST105. The display image in the display state is displayed on the display 50.

  Then, in step ST106, the user instruction information is continuously acquired, and in step ST107, it is determined whether or not the display state is changed. When it is determined that the display state is changed, the process returns to step ST102, and the changed display image is generated based on the user instruction. For example, when a user's instruction to change the display image such as enlargement or rotation processing is given to the three-dimensional image data with a mouse or a touch panel, a new display image is displayed in step ST102 based on the user instruction. Is generated. Then, a determination is made as to whether or not the individual can be specified for this new display image, and if it is determined that the individual can be specified, anonymization processing is performed for the display image at that time. .

  That is, the determination function 23 determines whether or not the individual of the subject can be identified from the display image every time the display state of the display image is changed in response to a user instruction. Then, when it is determined that the individual can be identified from the display image, the anonymization processing function 24 performs anonymization processing that prevents the individual from being identified from the display image each time the display state of the display image is changed. The display image after the anonymization process is displayed on the display 50 in real time.

  Here, the anonymization process is at least a masking process, a transparentizing process, a smoothing process, a deforming process, a hair removing process, and a hair increasing process for the epithelial region of the anatomical part of the display image or the three-dimensional image data. This is a process that makes it difficult to identify the individual of the subject by performing one process. Note that when these anonymization processes are performed on the three-dimensional image data, the three-dimensional image data is only temporarily processed and rendering processing is performed to generate a display image. It does not process the data permanently.

  Further, the anatomical part to be subjected to the anonymization process may be selected based on the size of the part-specific determination index value. For example, if the region-specific determination index values for “ears” and “nose” are smaller than the region-based determination index values for “eyes” and “mouths”, anonymization processing is performed for “eyes” and “mouths”. It is also conceivable that the anonymization process is not performed on the “ears” and the “nose”.

  Hereinafter, a specific example of the above operation will be described with reference to FIGS. 6 to 11. FIG. 6 is a diagram showing a first specific example. As shown in the upper left of FIG. 6, the display image of the first specific example is a volume rendering image in which the head of the subject facing forward is enlarged.

  In the lower center of FIG. 6, a determination table corresponding to the display image of the first specific example is illustrated. In the display image of the first specific example, all the “eyes”, “mouths”, “nose”, and “ears” are visible, so “25” points are set for all the basic values corresponding to these. Is assigned. Also, since the subject is facing the front, all parts are not hidden, and there is no reduction in area due to tilting, so all of the “eyes”, “mouths”, “nose”, and “ears” A weight of “1.0” is assigned to the area ratio weight W1 of.

  In addition, the "eyes", "mouths", "nose", and "ears" of this display image are completely opaque because they are not transparent. A weight of “1.0” is also assigned to all the transparency weights W2 of “”, “nose”, and “ear”.

  On the other hand, it is considered that this subject has strong individuality in the eyes and mouth, whereas the nose and ears have relatively common shapes, and the individuality is not so strong. Therefore, "1.0" is assigned to the "eyes" and "mouths" as the individuality weight W3, and "0.4" is assigned to the "nose" and "ears".

  As a result, the region-specific determination index values are calculated as “25”, “25”, “10”, and “10” for “eye”, “mouth”, “nose”, and “ear”, respectively. To be done. Further, the determination index value for each part is added to calculate the overall determination index value as "70".

  Since this comprehensive determination index value is larger than the predetermined threshold value “30.0”, it is determined that the anonymization process is performed.

  The display image in the upper right part of FIG. 6 exemplifies the display image after the anonymization process. In the display image in the upper right part of FIG. 6, anonymization processing is performed by performing masking processing that outlines the eyes, nose, and mouth.

FIG. 7 is a diagram showing a second specific example. As shown in the upper center of FIG. 7, the display image of the second specific example is a volume rendering image in which the head of the display image of the first specific example is rotated leftward by 45 degrees.

  In the lower center of FIG. 7, a determination table corresponding to the display image of the second specific example is illustrated. Also in the display image of the second specific example, at least a part of each of the “eye”, “mouth”, “nose”, and “ear” is visible, so all the basic values corresponding to each of these parts are displayed. , “25” points are allocated to.

  On the other hand, regarding the area ratio weight W1, the left ear is hidden because the subject is facing left by 45 degrees. Therefore, the area ratio weight W1 for the "ears" is set to "0.5". In addition, the area in which the "eyes", "mouths", and "nose" are visible is smaller than the area when facing the front because the subject faces 45 degrees to the left. Therefore, the area ratio weight W1 for the "eye", "mouth", and "nose" is set to "0.7", respectively. The same value is assigned to the transparency weight W2 and the individuality weight W3 as in the first specific example.

  As a result, the region-specific determination index values are “17.5”, “17.5”, “7.0”, and “eye”, “mouth”, “nose”, and “ear”, respectively. And “5.0”. Then, the determination index value for each part is added to calculate the overall determination index value as “47.0”.

  Since this comprehensive determination index value is larger than the threshold value “30.0”, it is determined that the anonymization process is performed.

  The display image in the upper right part of FIG. 7 exemplifies the display image after the anonymization process. In the display image in the upper right part of FIG. 7, similarly, the anonymization process is performed by performing the masking process for blanking the eyes, nose, and mouth similarly to the first specific example.

  FIG. 8 is a diagram showing a third specific example. As shown in the upper center of FIG. 8, the display image of the third specific example is a volume rendering image in which the head of the display image of the first specific example is rotated 90 degrees to the left.

  In the lower center of FIG. 8, a determination table corresponding to the display image of the third specific example is illustrated. Also in the display image of the third specific example, at least some of the “eyes”, “mouths”, “nose”, and “ears” are visible, so all the basic values corresponding to these respective parts , “25” points are allocated to.

  On the other hand, regarding the area ratio weight W1, the left eye and the left ear are hidden because the subject is facing leftward by 90 degrees. In addition, the area in which the "eyes", "mouths", and "nose" are visible is smaller than the area when facing the front because the subject is facing 90 degrees to the left. Therefore, the area ratio weights W1 for “eyes”, “mouths”, “nose”, and “ears” are set to “0.4,”, “0.4,”, “0.6”, and “0. 5, ”is set. The same value is assigned to the transparency weight W2 and the individuality weight W3 as in the first specific example.

  As a result, the region-based determination index values are “10.0”, “10.0”, “6.0”, and “eye”, “mouth”, “nose”, and “ear”, respectively. And “5.0”. Then, the determination index value for each part is added to calculate the comprehensive determination index value as “31.0”.

  Since this comprehensive determination index value is larger than the threshold value “30.0”, it is determined that the anonymization process is performed.

  The display image in the upper right part of FIG. 8 exemplifies the display image after the anonymization process. In the display image in the upper right part of FIG. 8, the anonymization process is performed by performing the masking process for blanking the eyes, nose, and mouth in the same manner as the first and second specific examples.

  FIG. 9 is a diagram showing a fourth specific example. As shown in the upper center of FIG. 9, the display image of the fourth specific example is a volume rendering image obtained by rotating the head of the display image of the first specific example leftward by 135 degrees.

  In the lower center of FIG. 9, a determination table corresponding to the display image of the fourth specific example is illustrated. In the display image of the fourth specific example, “eyes”, “mouths”, and “nose” are not visible at all, and only part of the right ear is visible. Therefore, "0", "0", "0" and "25" points are assigned to the respective basic values of "eye", "mouth", "nose" and "ear".

  As a result, the site-specific determination index values of “eye”, “mouth”, and “nose” are all “0.0” regardless of the weight value. On the other hand, for “ears”, the basic value “25 points” is multiplied by the area ratio weight W1 of “0.5”, the transparency weight W2 of “1.0”, and the individuality weight W3 of “0.4”, respectively. As a result, the determination index value for each part becomes “5.0”. The determination index value for each part is added to calculate the overall determination index value as "5.0". Since this comprehensive determination index value is smaller than the threshold value “30.0”, it is determined that the anonymization process is not performed.

  The display image in the upper right part of FIG. 9 is the same as the display image in the upper center part of FIG. 9, and this image that has not been anonymized is displayed on the display 50.

  FIG. 10 is a diagram showing a fifth specific example. As shown in the upper center of FIG. 10, in the display image of the fifth specific example, the head of the display image of the first specific example is rotated 180 degrees leftward, that is, the volume rendering of the head facing backward. It is an image.

  In the display image of the fifth specific example, like the display image of the fourth specific example, "eyes", "mouths", and "nose" are not visible at all, and part of the left and right ears are visible. Only. Therefore, "0", "0", "0" and "25" points are assigned to the respective basic values of "eye", "mouth", "nose" and "ear".

  As a result, the site-specific determination index values of “eye”, “mouth”, and “nose” are all “0.0” regardless of the weight value. On the other hand, for “ears”, the basic value “25 points” is multiplied by the area ratio weight W1 of “1.0”, the transparency weight W2 of “1.0”, and the individuality weight W3 of “0.4”, respectively. As a result, the region-specific determination index value becomes “10.0”. The determination index value for each part is added to calculate the overall determination index value as “10.0”. Since this comprehensive determination index value is smaller than the threshold value “30.0”, it is determined that the anonymization process is not performed.

  The display image in the upper right portion of FIG. 10 is the same image as the display image in the upper center of FIG. 10, and this image that has not been anonymized is displayed on the display 50.

  FIG. 11 is a diagram showing a sixth specific example. As shown in the upper center of FIG. 11, the display image of the sixth specific example is a volume rendering image enlarged so that only the left eye portion of the head of the display image of the first specific example is displayed. Is.

  In the display image of the sixth specific example, like the display image of the fourth specific example, the "mouth", "nose", and "ear" are not visible at all, and only the left eye is visible. Therefore, “0” points are assigned to the basic values of “mouth”, “nose” and “ear”, and “25” points are assigned only to the basic values of “eye”.

  As a result, the region-based determination index values of ““ mouth ”,“ nose ”and“ ear ”are all“ 0.0 ”regardless of the weight value. On the other hand, for the “eye”, the basic value “25 points” is multiplied by the area ratio weight W1 of “0.5”, the transparency weight W2 of “1.0”, and the individuality weight W3 of “1.0”, respectively. As a result, the region-specific determination index value becomes “12.5”. The determination index value for each part is added to calculate the overall determination index value as “12.5”. Since this comprehensive determination index value is smaller than the threshold value “30.0”, it is determined that the anonymization process is not performed.

  The display image in the upper right part of FIG. 11 is the same as the display image in the upper center of FIG. 11, and this image that has not been anonymized is displayed on the display 50.

  According to the medical image browsing apparatus 100 of the first embodiment described above, it is determined whether the individual of the subject can be identified from the display image, and if it is determined that the individual can be identified, the display image is compared with the display image. Anonymization processing is performed. Then, each time the display state of the display image is changed in response to a user's instruction such as rotation / movement or enlargement / reduction, it is determined in real time whether or not an individual can be identified, and anonymity is determined according to the determination result. The conversion process is supposed to be performed in real time.

  With such processing, when the medical image browsing apparatus 100 of the first embodiment is used to display image data of a certain subject through the display 50 for use in academic conference presentations, teaching materials of medical departments, etc. However, it is possible to prevent the individual from being specified from the displayed image. Further, the anonymization process does not permanently process the original image data, but processes the display image generated from the image data, or even when the image data is processed, generates the display image and displays it on the display 50. The image data is processed only during the period displayed in. Therefore, the diagnosis and treatment of the head such as eyes, nose, mouth, and ears will not be seriously affected. For example, in the display image of the first specific example (see FIG. 6) that displays the entire face, the eyes, nose, mouth, etc. are masked, but after that, by changing the display state of the display image by a user operation, For example, the anonymization process can be canceled easily by changing the display state of the display image of the sixth specific example (see FIG. 11) in which only the left eye is enlarged for eye diagnosis and treatment.

(Second embodiment)
FIG. 12 is a flowchart showing an operation example of the medical image browsing apparatus 100 according to the second embodiment. The difference from the first embodiment is that in determining whether an individual can be identified from a display image, the process of step ST200 is added instead of the method of using the determination table used in the first embodiment. It is a point to judge whether or not the individual of the subject can be specified from the displayed image by the face recognition technology including the artificial intelligence (AI) technology.

  For example, with respect to the display image described above, face recognition using artificial intelligence is performed, and an index value such as reliability indicating the certainty when identifying an individual for which artificial intelligence is determined is acquired, and the index value Whether or not to perform the anonymization process may be determined by the threshold value determination.

  Also in the medical image browsing apparatus 100 of the second embodiment, the same effect as that of the first embodiment can be obtained. The configuration example of the second embodiment is the same as the configuration example (block diagram shown in FIG. 2) of the first embodiment.

(Third Embodiment)
FIG. 13 is a flowchart showing an operation example of the medical image browsing apparatus 100 according to the third embodiment. The difference from the third embodiment is that step ST300 for determining whether anonymization processing is necessary is added before the processing of step ST103 for determining whether an individual can be identified from the display image. It is a point.

  For example, if a doctor in charge wants to view an image taken in the order of the doctor in charge for diagnosing a subject, the doctor in charge knows the face of the subject individually, and thus the anonymous No conversion processing is required. In such a case, it is determined that the anonymization process is unnecessary based on the identification information of the doctor in charge of capturing the image and the identification information of the user who is browsing (that is, the same doctor). You can

  Further, there is a case in which the anonymization process is unnecessary when the image data is used in the hospital where the image of the subject is imaged, and the anonymization process is necessary when the image data is used outside the hospital, such as at a conference presentation. In such a case, it is possible to determine whether the anonymization process is necessary based on the standard information of the image data. For example, when the standard information of the image data is the DICOM standard, use in a hospital is assumed, and therefore it is determined that the anonymization process is unnecessary. On the other hand, if the standard information of the image data is a general-purpose standard such as JPEG, it can be determined that the anonymization process is necessary.

  According to the medical image browsing apparatus 100 of the third embodiment, in addition to the effects of the first and second embodiments, whether or not the anonymization process is necessary can be easily determined.

(Other embodiments)
FIG. 14 is a diagram showing a configuration example of a medical image processing apparatus 530 in which the input device 40 such as a mouse and a keyboard and the display 50 are removed from the medical image browsing apparatus 100 described above. The medical image processing apparatus 530 is, so to speak, an apparatus corresponding to the configuration of the workstation main body. As can be seen from the configuration shown in FIG. 14, the configuration of the medical image processing apparatus 530 is substantially the same as that of the medical image browsing apparatus 100 described above, and a description thereof will be omitted. The medical image processing apparatus 530 also has the same effects as those of the above-described embodiments of the medical image browsing apparatus 100.

  FIG. 15 shows a configuration example in which the configuration of the medical image browsing apparatus 100 described above is incorporated into a medical image diagnostic apparatus 510 (that is, modality 510) such as an X-ray CT apparatus 511, an MRI apparatus 512, and an ultrasonic diagnostic apparatus 513. It is a figure. In general, the medical image diagnostic apparatus 510 has an imaging unit (that is, a scanner) 600 having a gantry and a probe, and a control unit. The control unit includes a processing circuit 20 that processes data captured by the image capturing unit 600 to generate an image, a display 50 that displays the generated image, and the like. As can be seen from the configuration shown in FIG. 15, the configuration of the medical image diagnostic apparatus 510 is substantially the same as that of the medical image browsing apparatus 100 described above, and a description thereof will be omitted. The medical image diagnostic apparatus 510 can also achieve the same effects as those of the above-described embodiments of the medical image browsing apparatus 100.

  In each of the above-described embodiments, the target of the determination as to whether or not the individual of the subject can be identified is mainly the “face (or head)”, and further, the elements used for the determination are “eyes”, “ Examples have been described using individual anatomical parts such as the "mouth," "nose," "ear."

  However, the determination target in each embodiment and the elements used for the determination are not limited to these, and a wider range of the subject can be the determination target, or conversely, more It is also possible to set a narrow range as a determination target.

  For example, the target of the determination as to whether or not the individual of the subject can be identified is the “entire upper body”, and the element used for the determination is one or more anatomical parts inside the upper body, for example, rib running. The shape or the head shape may be used.

  Further, for example, the target of the determination of whether the individual of the subject can be specified is "a specific site in the face", for example, "nose", and as an element used for the determination, an anatomical site in the nose, For example, a nasal cavity or a nasal bone may be used.

  According to at least one embodiment described above, it is possible to prevent an individual from being discriminated and specified from a medical image without impairing the original purpose of the medical image such as diagnosis and treatment.

  The image data acquisition function and the image processing function described in each embodiment are examples of the generation unit described in the claims. Further, the determination function and the anonymization processing function described in each embodiment are examples of the determination unit and the processing unit described in the claims.

  Although some embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the spirit of the invention. These embodiments and modifications thereof are included in the scope and the gist of the invention, and are also included in the invention described in the claims and the scope equivalent thereto.

20 Processing Circuit 21 Image Data Acquisition Function 22 Image Processing Function 23 Judgment Function 24 Anonymization Processing Function 30 Storage Circuit 40 Input Device 50 Display 100 Medical Image Viewing Device 510 Medical Image Diagnostic Device 520 Image Server 530 Medical Image Processing Device

Claims (13)

Display,
A generation unit that generates a display image to be displayed on the display from medical image data of a subject imaged, according to a user instruction,
Every time the display state of the display image is changed according to the user instruction, a determination unit that determines whether the individual of the subject can be specified from the display image,
When it is determined that the individual can be identified from the display image, anonymization processing for preventing the individual from being identified from the display image is performed in real time every time the display state of the display image is changed, and the anonymization is performed. A processing unit for displaying the display image after processing on the display,
A medical image browsing device including. The determination unit,
Detecting the anatomical site necessary to identify an individual from the display image,
For each of the anatomical parts, calculate a judgment index value for each part representing the ease of individual identification,
It is determined whether an individual can be identified from the display image using the determination index value for each part,
The medical image browsing apparatus according to claim 1. The determination unit,
Calculate the determination index value for each site by weighting the basic value assigned to each anatomical site,
By the threshold determination for the total determination index value obtained by adding the region-based determination index value, it is determined whether the individual can be identified from the display image,
The medical image browsing apparatus according to claim 2. The determination unit,
Weighting the basic value using the degree of exposure of the anatomical region in the display image, to calculate the region-based determination index value,
The medical image browsing apparatus according to claim 3. The degree of exposure of the anatomical region is the anatomical region visible in the display state of the display image with respect to the area of the anatomical region when the entire anatomical region is visible. Is the ratio of the area of the parts,
The medical image browsing apparatus according to claim 4. The degree of exposure of the anatomical region is the ratio of the area of the anatomical region visible in the display state of the display image to the area when the entire anatomical region is visible. , The product of the transparency assigned to the anatomical part,
The medical image browsing apparatus according to claim 4. The determination unit,
As the weight of the individuality indicating how different the shape of the anatomical part is from the standard shape, the part-based judgment index value is further weighted, and the total judgment index value is calculated.
The medical image browsing apparatus according to claim 4. The processing unit is
Based on the size of the region-specific determination index value representing the ease of identifying an individual, at least one of the display image and the image data is processed into a region corresponding to an anatomical region, and the anonymization is performed. Do the processing,
The medical image browsing apparatus according to claim 2. The anonymization process includes at least one process of a masking process, a clarification process, a smoothing process, a deforming process, a hair removing process, and a hair increasing process for the epithelial region of the anatomical site,
The medical image browsing apparatus according to claim 8. The determination unit determines whether or not the individual of the subject can be identified from the display image by a face recognition technique including an artificial intelligence technique,
The medical image browsing apparatus according to claim 1. The image data is two-dimensional or three-dimensional image data acquired by an X-ray CT (Computed Tomography) apparatus, an MRI (Magnetic Resonance Imaging) apparatus, or an ultrasonic diagnostic apparatus.
The medical image browsing apparatus according to any one of claims 1 to 10. A generation unit that generates a display image to be displayed on a display from medical image data obtained by imaging an object according to a user instruction,
Every time the display state of the display image is changed according to the user instruction, a determination unit that determines whether the individual of the subject can be specified from the display image,
When it is determined that the individual can be identified from the display image, anonymization processing for processing at least one of the display image and the image data so that the individual cannot be identified from the display image is performed. A processing unit that performs in real time each time the display state is changed,
A medical image processing apparatus comprising: An imaging unit for imaging the subject,
Display,
A generation unit that generates a display image to be displayed on the display from medical image data of a subject imaged, according to a user instruction,
Every time the display state of the display image is changed according to the user instruction, a determination unit that determines whether the individual of the subject can be specified from the display image,
When it is determined that the individual can be identified from the display image, anonymization processing for processing at least one of the display image and the image data so that the individual cannot be identified from the display image is performed. Performed in real time every time the display state is changed, and display the display image after the anonymization process on the display, a processing unit,
A medical image diagnostic apparatus comprising:

Images