JP7332362B2 - Medical image processing apparatus, medical image processing system, and medical image processing method - Google Patents

Description

TECHNICAL FIELD Embodiments of the present invention relate to a medical image processing apparatus, a medical image processing system, and a medical image processing method.

Today, various diseases can be diagnosed by analyzing image data generated by imaging a subject with a modality such as an X-ray CT (Computed Tomography) device, an MRI (Magnetic Resonance Imaging) device, or an ultrasonic diagnostic device. is being done.

For example, X-ray images and X-ray CT images are used to diagnose diffuse lung disease. Diffuse lung disease is a general term for lung diseases in which lesions spread relatively evenly in the left and right lungs, and includes various diseases. Interstitial pneumonia is a typical diffuse lung disease, but there are other diffuse lung diseases caused by infections, tumors, and the like. In addition, interstitial pneumonia is not a single disease. In addition to idiopathic interstitial pneumonia, which is the most common, connective tissue disease, hypersensitivity pneumonitis, pneumoconiosis, and occupational lung disease are also included in the broad definition of interstitial pneumonia. included. Diffuse lung disease can thus be classified into several distinct disease types, and these distinct disease types may hereinafter be referred to as disease causes.

On the other hand, X-ray CT images of diffuse lung disease are known to be classified into several types of texture patterns. In addition, a method has been proposed in which texture analysis is performed on lung X-ray CT images to classify the lung field into several texture patterns, and the volume ratio of each texture pattern to the entire lung field and its change over time are displayed. ing.

However, with these conventional techniques, it is difficult to grasp changes in the disease state in a local region of tissue, such as whether the local region of tissue such as the lung is in the direction of recovery or in the direction of exacerbation. I couldn't do it.

U.S. Patent Application Publication No. 2014/0184608

Uchiyama Y, Katsuragawa S, Abe H, et al. Quantitative computerized analysis of diffuse lung disease in high-resolution computed tomography. Med Phys. AAPM; 2003;30(9):2440－2454.

The problem to be solved by the present invention is to make it possible to easily grasp changes in disease states in localized regions of tissues from medical images.

A medical image processing apparatus according to an embodiment includes an acquisition unit, an analysis unit, and an estimation unit. The acquiring unit acquires a plurality of images obtained by imaging the same subject at different imaging times. The analysis unit classifies the tissue properties of the subject into a plurality of tissue property classes by analyzing the tissue properties of the subject based on the pixel values of the respective regions of the plurality of images, and classifying the classified tissue properties of the subject. A tissue attribute class is assigned to each region of the plurality of images. The estimating unit estimates changes in the disease state of the subject from changes in the tissue attribute class in corresponding regions of the plurality of images.

1 is a diagram showing a configuration example of a medical image diagnostic system including a medical image processing apparatus according to a first embodiment; FIG. 1 is a block diagram showing a configuration example of a medical image processing apparatus according to a first embodiment; FIG. 4 is a flowchart showing an operation example of the medical image processing apparatus of the first embodiment; 4A and 4B are schematic diagrams showing an example of a first image and a second image; FIG. FIG. 4 is a view for explaining the processing concept of steps ST102 to ST105 in the flowchart of FIG. 3; 4A and 4B are diagrams showing examples of different types of texture patterns; FIG. FIG. 4 is a diagram showing an example in which left and right lung regions are classified into a plurality of texture pattern types; The figure which shows an example of the lookup table for estimation of a state change direction. The figure which shows an example of a disease state change map. 6 is a flow chart showing an example of the operation of the medical image processing apparatus according to the modified example of the first embodiment; FIG. 11 shows an example of a lookup table for estimating the direction and speed of change in disease state. FIG. 4 is a diagram showing an example of a disease state change map that depicts the direction and speed of disease state change. 8 is a flowchart showing an operation example of the medical image processing apparatus according to the second embodiment; FIG. 4 is a diagram showing an example of a lookup table for estimating the state change direction for each disease cause; 10 is a flowchart showing an operation example of a medical image processing apparatus according to a modified example of the second embodiment; FIG. 10 is a diagram for explaining the operation concept of processing for estimating the cause of a disease from changes in the type of texture pattern;

A medical image processing apparatus according to an embodiment will be described with reference to the accompanying drawings. It should be noted that, in the following embodiments, portions denoted by the same reference numerals operate in the same manner, and overlapping descriptions will be omitted as appropriate.

FIG. 1 is a diagram showing a configuration example of a medical image processing system including a medical image processing apparatus 100 according to an embodiment. A medical image processing system is a series of medical image processing systems that acquire, process, store, and use medical images, for example, in a hospital.

The medical image processing system includes a modality 510 (that is, a medical image diagnostic device 510) such as an X-ray CT device 511, an MRI device 512, and an ultrasonic diagnostic device 513 for acquiring medical images from a subject such as a patient, as well as an image server. and the medical image processing apparatus 100 . These devices are connected to each other via, for example, a hospital network 500 so as to exchange various data and medical images.

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

The input interface circuit 10 is an interface circuit for inputting data via a wired or wireless network, a dedicated or general-purpose communication line, or inputting data via a storage medium such as an optical disk or a USB memory. . In the medical image processing apparatus 100 of the first embodiment, through the input interface circuit 10, a first image and a second image captured by a modality 510 such as an X-ray CT apparatus 511, or stored in an image server. A first image and a second image are obtained.

Here, the first image and the second image are images of the same subject taken on different dates and times. For example, the second image is an image captured on a date and time later than the date on which the first image was captured. The first image and the second image will be described in more detail later.

The storage circuit 30 is a storage medium including a ROM (Read Only Memory), a RAM (Random Access Memory), and an external storage device such as an HDD (Hard Disk Drive) and an optical disk device. The storage circuit 30 stores various information and data including a lookup table 31 to be described later, and also stores various programs executed by the processor included in the processing circuit 20 .

The input device 40 is, for example, a mouse, keyboard, trackball, touch panel, etc., 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, plasma display panel, or organic EL panel.

The processing circuit 20 is, for example, a circuit including a CPU or a dedicated or general-purpose processor. The processor implements various functions described later 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). These hardware can also realize various functions described later. The processing circuit 20 can also implement 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. Further, when a plurality of processors are provided, a storage medium for storing programs may be provided individually for each processor, or a single storage medium may collectively store programs corresponding to the functions of all processors. good too.

(Operation of the first embodiment)
The processing circuit 20 of the first embodiment has the functions shown in FIG. Implement function 25. 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 acquires a plurality of images obtained by imaging the same subject at different imaging times. For example, a first image captured at a first date and time and a second image captured at a second date and time later than the first date and time are acquired. Although the first image and the second image are not particularly limited, they are, for example, an X-ray image of lungs or an X-ray CT image.

The texture analysis function 22 classifies each region of the plurality of images acquired by the image data acquisition function 21 into a plurality of texture pattern types. For example, the texture analysis function 22 performs a known texture analysis on the first image and the second image acquired by the image data acquisition function 21, and divides each region of the first image and the second image into a plurality of Categorized into types of texture patterns.

The disease state change estimating function 23 estimates changes in the disease state of the subject from changes in texture pattern types between corresponding regions of the plurality of images classified by the texture analysis function 22 . For example, the disease state change estimation function 23 determines the direction of change in the disease state of the subject in the local region based on the change in the type of texture pattern in the corresponding local regions of the first image and the second image. , b) direction of exacerbation, and/or c) no change. Note that the corresponding region (that is, corresponding region) or corresponding local region means, for example, an anatomically corresponding region (or local region) between a plurality of images.

The disease state change map generation function 24 generates a map image in which pixel values that enable identification of disease state changes are assigned to each corresponding region. More specifically, the disease state change map generation function 24 draws the direction of change in the disease state estimated by the disease state change estimation function 23 in units of pixels or in units of pixel groups composed of two or more pixels. generate a modified disease state change map. For example, the disease state change map generation function 24 may determine that the direction of disease state change is different, including at least one of different colors, different densities, different numbers, and different symbols, for each pixel or each group of pixels. According to an aspect, a disease state change map is generated to be identifiably delineated.

The display control function 25 causes the display 50 to display the generated disease state change map, for example, according to a user's instruction.

FIG. 3 is a flow chart showing an operation example of the medical image processing apparatus 100 of the first embodiment. 4 to 9 are diagrams for explaining the operation concept of the medical image processing apparatus 100 of the first embodiment. Hereinafter, the operation of the medical image processing apparatus 100 will be described more specifically along the flow of FIG. 3 and with reference to FIGS. 4 to 9. FIG.

First, the first image is acquired in step ST100 of FIG. 3, and the second image is acquired in step ST101. FIG. 4 is a schematic diagram showing an example of the first image and the second image. Although the first image and the second image are images of the same subject, they are captured at different times. The first image is, for example, an image captured in Examination 1 performed on May 8, 2018, and the second image is an image captured in Examination 2 performed after Examination 1, for example, in August 2018. It is an image captured in Examination 2 performed on the 8th.

In the above example, the interval between the date when the first image was captured and the date when the second image was captured is 3 months, but this is just an example, and may be 1 week, 1 month, 6 months, 12 months, etc. , is determined by a physician's judgment. In addition, the number of times of imaging is not limited to two, and two desired images on different imaging dates are selected from a plurality of images taken on a plurality of imaging dates of three or more times, and two images on earlier imaging dates are selected. One image may be the first image and the other image may be the second image.

Note that the first and second images shown in FIG. 4 are merely schematic diagrams for explaining the operation of the medical image processing apparatus 100 of this embodiment. The first and second images are, for example, two-dimensional images obtained by cutting out a three-dimensional image of the lung field of the subject with the X-ray CT apparatus 511 at a certain coronal cross section.

As described above, X-ray images and X-ray CT images are used to diagnose, for example, diffuse lung disease. Diffuse lung disease is mainly treated with drugs such as steroids and immunosuppressants. In order to determine the therapeutic effect, it is extremely important to monitor the shadow pattern of X-ray CT images, that is, the change in texture pattern over time.

On the other hand, it is known that X-ray CT images of lesions of diffuse lung disease are classified into several types of texture patterns. It is also known that the type of texture pattern changes as the disease progresses from the initial state to the exacerbation direction. Conversely, it is also known that the type of texture pattern changes when the therapeutic effect increases and the disease goes from an exacerbated state to a recovery direction.

FIG. 4 schematically shows the temporal change of such texture patterns in the first image and the second image. The dark regions in the background (including the periphery) of the lung fields in each of the first and second images indicate normal, disease-free regions. On the other hand, in FIG. 4, a plurality of light-colored hatched areas are illustrated in addition to the dark-background area. In FIG. 4, these light-colored hatched areas correspond to diseased areas. Also, different hatching patterns in FIG. 4 correspond to different shadow patterns (texture patterns) in the X-ray CT image.

In the example shown in FIG. 4, of the left and right lungs of the subject, the upper part of the right lung (the left lung in FIG. 4) has the same type of light-colored texture pattern between the first image and the second image. However, the area of the texture pattern in the right lung of the second image is larger than the area of the texture pattern in the right lung of the first image. This indicates that the number of right lung lung lesions was higher as of August 8, 2018, when the second image was obtained (three months after the first image was obtained) than as of May 8, 2018, when the first image was obtained. In the right lung, the disease area is expanding, which means that the disease is in the process of exacerbation.

On the other hand, regarding the left lung (the lung on the right side in FIG. 4), in the first image, a texture pattern of a type different from that of the right lung occurs over a wide range. The area has shrunk in the second image. On the other hand, in the second image, a texture pattern different from the above two types occurs in the lower part of the left lung. This means that areas in the direction of recovery and areas in the direction of deterioration coexist in the left lung.

Conventionally, changes in texture patterns in such X-ray CT images have often been subjectively evaluated by doctors. On the other hand, the medical image processing apparatus 100 of the present embodiment detects local changes in the type of texture pattern and aggravates the disease from the local change in the type of texture pattern, as described below. It enables the device to objectively estimate the direction of change in the disease state, whether it is moving in the direction of recovery or moving in the direction of recovery.

Returning to FIG. 3, in step ST102, texture analysis is performed on the obtained first image, and in step ST103, the regions of the first image are classified into a plurality of texture pattern types based on the texture analysis. Similarly, in step ST104, texture analysis of the obtained second image is performed, and in step ST105, regions of the second image are classified into a plurality of texture pattern types based on the texture analysis.

FIG. 5 is a diagram for explaining the concept of processing from steps ST102 to ST105. A rectangular area surrounded by a broken line shown in the second image of FIG. 5 is a determination window for texture analysis. In texture analysis, for example, the feature amount of the target region is calculated by analyzing the pixel values within this determination window (or the CT value for each pixel if the image to be analyzed is an X-ray CT image). For example, statistics such as the average value and standard deviation of pixel values, histogram distribution of pixel values, parameters related to spectrum distribution, and the like are calculated as feature amounts.

By shifting the determination window in the left, right, up, and down directions as illustrated in FIG. 5, for example, the distribution of the feature amount of the entire image can be calculated. The shift step width may be in units of pixels or in units of multiple pixels. Alternatively, the entire image may be divided in advance into meshes each having the size of the determination window, and the feature amount in each mesh may be calculated sequentially or in parallel.

Further, in steps ST103 and ST105, the regions of the first and second images are classified into a plurality of texture pattern types based on the calculated feature amount. For example, as disclosed in Non-Patent Document 1, medical images (for example, X-ray CT images) of diseased areas such as diffuse lung disease are classified into multiple types of texture patterns.

For example, in steps ST103 and ST105, as shown in FIG. 6, the first image and the second image are divided into normal texture patterns (type "A") and five types of abnormal texture patterns (type "B", type "C", type "D", type "E", and type "F").

For example, the texture pattern is a type “A” “normal” pattern, a type “B” “ground-glass” pattern (ground-glass opacities), a type “C” “reticular and linear opacities” pattern. ), type 'D' 'nodular opacities', type 'E' 'honeycombing' pattern, type 'F' 'consolidation' pattern, etc. do.

Each texture pattern in FIG. 6 is merely an example for explaining the operation of this embodiment, and is not intended to be medical correctness or accuracy.

FIG. 7 is a diagram showing an example in which the left and right lung regions of the first image and the second image are classified into a plurality of texture pattern types by the processing in steps ST103 and ST105.

In the example shown in the first image, the upper part of the right lung is classified with a texture pattern of type 'B', which indicates a 'frosted glass' pattern, and the rest of the right lung is classified as type 'A', which indicates a 'normal' pattern. are classified as texture patterns. Also, most of the left lung in the first image is classified as a texture pattern indicating a type "C" "reticular" pattern.

On the other hand, in the example shown in the second image, a significant portion of the right lung is classified as a texture pattern of type "B" indicating a "frosted glass" pattern. The left lung of the second image is classified into a texture pattern indicating a type "C" "reticulated" pattern and a texture pattern indicating a type "D" "nodular" pattern.

In the example shown in FIG. 7, a square mesh of pixel groups (eg, 5×5 pixel groups) classifies the left and right lung regions into respective types of texture patterns. However, as described above, it is also possible to calculate the distribution of the feature quantity in pixel units by shifting the determination window of a predetermined size in the horizontal and vertical directions in pixel units. In this case, it is also possible to obtain the distribution of types of texture patterns in terms of pixel-by-pixel smoothness.

Returning to FIG. 3, in step ST106, the first image and the second image are aligned as necessary, and then the texture pattern changes in the corresponding local regions of the first image and the second image. , the direction of change in disease state is estimated.

Here, "estimating the direction of change in the disease state" means that the disease state, that is, the disease state of the diseased area, is in at least one of a) recovery direction, b) exacerbation direction, and c) no change. It is to presume that there is For example, the transition of the type of texture pattern and the direction of change in the disease state (that is, whether the diseased area is in the direction of recovery, in the direction of exacerbation, or in which direction no change is observed) By referring to the lookup table 31 associated with , the direction of change in disease state can be estimated. This lookup table 31 is stored in the storage circuit 30, for example. The disease state change estimating function 23 of the processing circuit 20 reads this lookup table 31 from the storage circuit 30 and uses it for the processing of step ST106.

An example of the lookup table 31 for estimating the state change direction is shown in the upper portion of FIG. In addition, the direction of change in the disease state and the transition of texture pattern types are shown in the lower part of FIG.

The diagram at the bottom of FIG. 8 shows that as the type of texture pattern of the affected area transitions from 'A' to 'B', from 'B' to 'C', and from 'C' to 'D', the disease state changes. , indicates a change from a normal state to a direction in which symptoms become severe (that is, to aggravation). Conversely, in the diagram at the bottom of FIG. 8, the type of texture pattern of the affected area transitions from 'D' to 'C', from 'C' to 'B', and from 'B' to 'A'. As the disease progresses, the disease state changes from a severe state to a normal state (that is, toward recovery).

The relationship between the direction of change in the disease state and the transition of texture pattern types shown in the lower part of FIG. It can be acquired from a database such as accumulation of progress information.

The lookup table in the upper part of FIG. 8 can be generated from the relationship between the direction of change in the disease state and the transition of texture pattern type shown in the lower part of FIG.

For example, in the lookup table of FIG. Types "A", "B", "C", and "D" of the texture patterns of the image (second image) captured in the examination 2 performed after the examination 1 of the specimen indicate changes in the disease state. They are related to each other by the three indicators shown (ie "no change", "worsening" and "recovery").

By referring to this lookup table, the direction of change in the disease state of the subject in the local region can be a) recovered from the change in the type of texture pattern in the local region corresponding to each other in the first image and the second image. direction, b) direction of exacerbation, and/or c) no change. For example, when a local region whose texture pattern type is “C” in the first image changes to texture pattern type “A” or “B” in the second image, the local region is restored. On the other hand, if the texture pattern changes to type "D", it is estimated that the local region is in the direction of aggravation.

As described above, diffuse lung disease is known to have multiple disease causes, and the relationship between the transition of the texture pattern type and the direction of change in the disease state may differ for each disease cause. have a nature. Therefore, in order to show that it corresponds to diffuse lung disease caused by a specific disease cause (for example, idiopathic interstitial pneumonia), the lookup table shown in FIG. (Cause of disease (a) in the example of FIG. 8) is associated.

In addition, in the lookup table of FIG. 8, one of the three symbols "0", "+" and "-" is assigned to each of the three indicators of "no change", "exacerbation" and "recovery". It is written together. These three symbols "0", "+" and "-" are for easy understanding of the correspondence with the "disease state change map" illustrated in FIG. It is not necessary to include the information of these three symbols '0', '+' and '-'.

Returning to FIG. 3, in step ST107, a disease state change map that depicts changes in disease state in local regions is generated. The process of step ST107 is a process corresponding to the disease state change map generation function 24 of FIG.

FIG. 9 is a diagram showing an example of a disease state change map. The disease state change map shown in FIG. 9 is divided into two images after being classified into types of texture patterns such as "A", "B", "C", and "D", that is, the first image shown in FIG. and the second image and the lookup table shown in FIG.

In FIG. 9 , the local regions drawn with the “+” symbol indicate local regions that are presumed to have deteriorated during the period from the imaging date of Examination 1 to the imaging date of Examination 2, and the “−” symbol Conversely, the local region depicted in ( ) indicates a local region that is estimated to have recovered during the period from the imaging date of Examination 1 to the imaging date of Examination 2 . On the other hand, the local regions drawn with the symbol “0” indicate local regions in which the disease state is estimated to have not changed during the period from the imaging date of Examination 1 to the imaging date of Examination 2 .

As for the resolution of the disease state change map, like the first and second images after texture pattern type classification shown in FIG. It may be rendered at a resolution per configured pixel group.

In addition, in the disease state change map illustrated in FIG. 9, the direction of change in disease state is identified by different symbols such as "+", "-", and "0", but it is not limited to this. do not have. For example, a disease state change map can be generated by identifiably delineating the direction of change in disease state by different aspects including at least one of different colors, different densities, different numbers, and different symbols. can.

The generated disease state change map is displayed on the display 50 provided in the medical image processing apparatus 100 in step ST108. Alternatively, the generated disease state change map may be transmitted to the modality 510 such as the X-ray CT apparatus 511 via the network 500 and displayed on the display of the modality 510 .

According to the medical image processing apparatus 100 of the first embodiment described above, the local region of the tissue such as the lung is in the direction of recovery or in the direction of deterioration. Changes in disease status in the area can be easily grasped.

(Modification of the first embodiment)
FIG. 10 is a flow chart showing an operation example of the medical image processing apparatus 100 according to the modification of the first embodiment. The difference from the first embodiment is the processing of step ST200, and other processing is the same as that of the first embodiment.

In step ST200, not only the direction of change in the disease state but also the speed of change in the disease state are further estimated from the change in texture pattern types in the local regions corresponding to each other in the first image and the second image. That is, the rate of exacerbation, such as whether the disease in the relevant tissue is gradually exacerbating or rapidly exacerbating, and conversely, whether the disease in the relevant tissue is gradually recovering or rapidly recovering. Estimate the recovery speed, such as whether or not

In the modification of the first embodiment, in order to estimate the rate of change in the disease state, a lookup table (FIG. 11) obtained by adding information on the change rate to the lookup table (FIG. 8) for estimating the state change direction. ) is used.

In the lookup table shown in FIG. 11, for example, if the texture pattern type "A" in the first image of the inspection 1 remains as the type "A" in the second image, "no change", type "B" If it transitions to type "C", it is determined that it is in the direction of "exacerbation" and the rate of exacerbation is "small", and if it transitions to type "C", it is determined that it is in the direction of "exacerbation" and the rate of exacerbation is "medium", In the case of transition to type "D", it is determined that the direction is "aggravation" and the rate of aggravation is "high".

Further, for example, if the texture pattern type "D" in the first image of the examination 1 remains as the type "D" in the second image, "no change", and if it transitions to the type "C", "recovery direction and the recovery speed is "small", and when the transition to type "B" is made, it is determined that the direction is "recovery" and the recovery speed is "medium", and the transition is made to type "A". In the case, it is determined that the direction is "recovery" and the recovery speed is "high".

Thus, the rate of disease state change in the lookup table shown in FIG. 11 is determined based on the transition distance between the types of texture patterns shown at the bottom of FIG.

In the lookup table shown in FIG. 11, the velocity in the direction of exacerbation is represented by numbers with positive signs "+1", "+2", and "+3" in order from the smallest to the largest, and the velocity in the recovery direction is , in order from the smallest to the largest, represented by numbers with negative signs “-1”, “-2”, and “-3”.

The symbols such as "+", "-" and "0" in the disease state change map shown in FIG. By substituting signed numbers such as 3' and '0', a disease state change map can be generated that depicts the rate of change in addition to the direction of change in disease state.

FIG. 12 is a diagram showing an example of a disease state change map that depicts the direction and speed of disease state change using identification information other than symbols and numbers. In the disease state change map shown in FIG. 12, the direction and speed of change in the disease state are depicted by the type of hatching pattern and the shade of hatching.

According to the modified example of the first embodiment, in addition to the direction of change in the disease state such as “aggravation”, “recovery”, and “no change”, the exacerbation speed and recovery speed are depicted on the disease state change map. , an objective and highly accurate diagnosis is possible.

In the above description, the rate of disease state change, such as exacerbation rate and recovery rate, was obtained based on the mutual transition distance between texture pattern types. The rate of disease state change can also be estimated based on the length of time between the capture date of the second image at . Furthermore, based on both the transition distance of texture pattern types from each other and the length of the period between the capture date of the first image in exam 1 and the capture date of the second image in exam 2, the disease state change It is also possible to estimate the velocity of

For example, when the interval between the imaging date of the first image and the imaging date of the second image is three months, the texture pattern type “A” in the first image of the examination 1 transitions to the type “B”. is in the direction of "exacerbation" and the rate of exacerbation is "small", but when the interval between the imaging date of the first image and the imaging date of the second image is one month, the first image of examination 1 Even if the type "A" of the texture pattern in Fig. 2 is changed to the type "B", it is determined that the exacerbation speed is not "low" but "medium" or "large".

(Second embodiment)
FIG. 13 is a flow chart showing an operation example of the medical image processing apparatus 100 of the second embodiment. As described above, diffuse lung disease is known to have multiple disease causes, and the relationship between the transition of texture pattern types and the direction of change in disease state may differ for each disease cause. be. Therefore, the medical image processing apparatus 100 of the second embodiment has a lookup table for estimating the state change direction for each disease cause.

In the second embodiment, steps ST300 and ST301 in FIG. 13 are different from the first embodiment. In step ST300, the disease cause designated by a user such as a doctor is obtained. For example, using the input device 40 such as a mouse or a keyboard, the user can identify the cause of diffuse lung disease as disease cause (a) (e.g., idiopathic interstitial pneumonia), disease cause (b) (e.g., Collagen disease), disease cause (c) (for example, hypersensitivity pneumonitis), disease cause (d) (for example, pneumoconiosis, occupational lung disease), etc., and the disease state change estimation function 23 of the processing circuit 20 specifies Get the cause of the disease.

In step ST301, a change in disease state is determined from a change in texture pattern type in local regions corresponding to each other in the first image and the second image, and a lookup table for estimating the state change direction corresponding to the specified disease cause. Estimate the direction of

FIG. 14 shows an example of four lookup tables for estimating the direction of state change that are different for each disease cause. As can be seen from the lookup table corresponding to the cause of disease (A) in the upper left of FIG. I'm assuming D.

On the other hand, as can be seen from the lookup table corresponding to disease cause (b) in the lower left of FIG. ”, and “F”. Also, as can be seen from each lookup table on the right side of FIG. 14, "A", "B", "H", and "J" are assumed as the types of disease texture patterns corresponding to disease cause (c). As types of disease texture patterns corresponding to disease cause (d), "A", "B", "K", and "F" are assumed.

In this way, it is believed that the types of assumed texture patterns differ depending on the types of disease causes. Therefore, when estimating the direction and speed of change in the disease state from the change in the type of texture pattern, it is preferable to consider the type of cause of the disease.

In the medical image processing apparatus 100 of the second embodiment, a lookup table for each cause of disease corresponding to a cause of disease specified by a user such as a doctor is referred from among a plurality of lookup tables for each cause of disease. estimates the direction of change in the subject's disease state. As a result, the direction of change in disease state can be estimated with high accuracy, which contributes to highly reliable diagnosis.

(Modification of Second Embodiment)
FIG. 15 is a flow chart showing an operation example of the medical image processing apparatus 100 according to the modification of the second embodiment. In the operation of the modified example of the second embodiment, the process in which step ST400 is added to the process of the second embodiment (the flow chart shown in FIG. 13) is performed. In step ST400, the cause of the disease is estimated from the change in texture pattern type.

FIG. 16 is a diagram explaining the concept of the process of step ST400. The central part of FIG. 16 illustrates a lookup table for disease cause estimation. This lookup table is substantially the same as the four lookup tables shown in FIG. 14, and is mutually convertible.

As described above, the types of texture patterns that may change with disease progression may differ depending on the cause of the disease, and not all types of texture patterns that may change with disease progression are the same. For example, as shown in the lookup table for disease cause estimation in FIG. Although it can occur, the change from texture pattern type 'B' to 'H' can only be caused by disease cause (c).

Therefore, in the modified example of the second embodiment, the cause of disease is estimated from the direction of change of the texture pattern used in the process of step ST301 and the lookup table for estimating the cause of disease. For example, as shown in the left part of FIG. 16, the change in the texture pattern of the subject (for example, the patient's back) changes from type "A" to type "B" in a certain local region, and changes to type "B" in another local region. Suppose that it changes from "B" to "Type H". In this case, the process of step ST400 presumes that the disease cause of patient A is disease cause (c).

As described above, in the modified example of the second embodiment, it is possible to estimate the cause of a disease from the change in the type of texture pattern, so that it is possible to provide the user with more useful diagnostic information. Become.

8, 11, 14, and 16 have exemplified some lookup tables, but the contents of these lookup tables are created, edited, and edited by the user. Alternatively, it is configured to be changeable.

(Other embodiments)
Up to this point, the texture analysis is mainly performed based on the pixel values of each region of a plurality of images (for example, the first image and the second image described above), so that each region is selected from one of a plurality of texture patterns. Embodiments have been described that classify However, embodiments of the present invention are not limited to these. Classifying the tissue properties of the specimen into a plurality of tissue properties classes, b) assigning the classified tissue properties classes to respective regions of the plurality of images, and c) changing the tissue properties class in respective corresponding regions of the plurality of images. estimating a change in a subject's disease state from

For example, as an example of analysis of tissue properties of a subject, in addition to the texture analysis described above, there is TIC (Time-Intensity Curve) analysis (for example, analysis of temporal change in contrast density). In this embodiment, TIC analysis is performed, TIC curves obtained from the analysis are classified into a plurality of types, ie, tissue characterization classes, and changes in the disease state of the subject are estimated from changes in TIC curve types. For example, when contrast-enhanced CT is performed on a subject with hepatocellular carcinoma in a normal liver, the TIC curve in the early phase (arterial phase) rises faster than the TIC curve in normal tissue, and the TIC curve in the late phase (equilibrium phase) rises earlier. ), wash out and the curve falls. By classifying the tissue pattern based on the degree and timing of this rise and fall, it is possible to estimate the transition whether each tissue is approaching (aggravating) the pattern of cancer from normal tissue or vice versa. can do.

Further, for example, an example of analysis of the tissue characterization of a subject includes an example of analyzing luminance values within an image region. In this embodiment, by analyzing the luminance values, for example, a histogram of luminance values is calculated and the shape of the histogram is classified into a plurality of types, ie, tissue attribute classes. For example, the histogram shape can be classified into multiple types using parameters such as kurtosis, which represents the sharpness of the histogram peaks, the number of histogram peaks, and the luminance value corresponding to the histogram peaks. A change in the subject's disease state is inferred from the change. For example, when CT imaging is performed on a subject having a pulmonary nodule, the luminance value distribution, that is, the histogram, is different in the area of the pulmonary nodule because the tissue density is different from that in the normal area. In the previous nodule example, the proportion of the low-density portion decreases while the proportion of the high-density portion increases compared to normal tissue. By classifying the tissue pattern based on the shape of this histogram, it is possible to estimate the change in whether each tissue is approaching (aggravating) the pattern of normal tissue or nodule.

Further, for example, as an example of analysis of tissue properties of a subject, there is an example of analyzing a plurality of types of index values within an image region. The multiple types of index values are, for example, luminance values and function index values in the region. Also, the function index value is an index value indicating the function of a tissue, for example, obtained from wall motion analysis or from cardiac nuclear medicine examination such as myocardial scintigraphy. In this embodiment, for example, regions are classified into a plurality of groups based on two values, a luminance value and a function index value. For example, a group with a high luminance value and a high function index value is designated as "group 1", a group with a high luminance value and a low function index value is designated as "group 2", and a group with a low luminance value and a low function index value is designated as "group 1". Group 3”, and the change in the subject's disease state is estimated from the change in the classified group.

According to at least one embodiment described above, it is possible to easily comprehend a change in disease state in a localized region of tissue from a medical image.

In the embodiments and modifications described so far, the functions 21 to 25 illustrated in FIG. 2 have been described as being realized by one medical image processing apparatus 100, but the present invention is not limited to this. . For example, a medical image processing system (not shown) in which a plurality of medical image processing apparatuses 100 are connected via a network may perform distributed processing of the functions 21 to 25 illustrated in FIG. For example, the first and second medical image processing apparatuses 100A and 100B constitute a medical image processing system, and some of the functions 21 to 25 shown in FIG. , and the remaining functions may be processed by the second medical image processing apparatus 100B. Since the plurality of medical image processing apparatuses 100 in the medical image processing system are connected to each other via a network, each medical image processing apparatus 100 can be installed at different locations.

Note that the storage circuit, image data acquisition function, texture analysis function, disease state change estimation function, disease state change map generation function, and display control function in the description of each embodiment are the storage circuits in the claims. It is an example of a unit, an acquisition unit, an analysis unit, an estimation unit, a generation unit, and a display control unit.

It should be noted that although several embodiments of the invention have been described, these embodiments are provided by way of example 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 modifications can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the scope of the invention described in the claims and equivalents thereof.

10 Input interface circuit 20 Processing circuit 21 Image data acquisition function 22 Texture analysis function 23 Disease state change estimation function 24 Disease state change map generation function 25 Display control function 30 Storage circuit 31 Lookup table 40 Input device 50 Display 100 Medical image processing apparatus 500 network 511 X-ray CT device 512 MRI device 513 ultrasonic diagnostic device

Claims (14)

an acquisition unit that acquires a plurality of images obtained by imaging the same subject at different imaging times;
By analyzing the tissue properties of the subject based on the pixel values of the respective regions of the plurality of images, the tissue properties of the subject are classified into a plurality of tissue property classes, and the classified tissue property classes are classified into , an analysis unit that allocates to each region of the plurality of images;
an estimating unit that estimates changes in the disease state of the subject from changes in the tissue attribute class in corresponding regions of the plurality of images ,
The analysis unit performs texture analysis based on pixel values of each region of the plurality of images, classifies each region into one of a plurality of texture patterns, and classifies the classified texture patterns as the assigned to each region of a plurality of images,
The estimating unit estimates the change in the disease state in the corresponding region and the rate of change in the disease state from the change in the texture pattern in the corresponding region of each of the plurality of images.
Medical image processing equipment. a generating unit that generates a map image in any of the plurality of images in which a pixel value that enables identification of a change in the disease state is assigned to each of the corresponding regions;
The medical image processing apparatus according to claim 1 . The image is an image of the lung field of the subject,
The texture pattern includes at least one of ground glass, net, nodule, honeycomb, and infiltration,
The medical image processing apparatus according to claim 1 or 2 . The acquisition unit acquires a first image captured at a first date and time and a second image captured at a second date and time later than the first date and time,
The analysis unit classifies each region of the first image and the second image into types of the plurality of texture patterns,
The estimating unit determines, based on changes in types of texture patterns in local regions corresponding to each other in the first image and the second image, that a direction of change in the disease state of the subject in the local region is a) a recovery direction , b) direction of exacerbation, and c) no change,
The medical image processing apparatus according to any one of claims 1 to 3 . The estimating unit estimates the direction of change in the disease state in units of pixels or in units of pixel groups composed of two or more pixels.
The medical image processing apparatus according to claim 4 . a display;
a generating unit for generating a disease state change map in which the direction of change in the disease state estimated by the estimating unit is drawn in units of pixels or in units of pixel groups composed of two or more pixels;
a display control unit for displaying the disease state change map on the display;
further comprising
The medical image processing apparatus according to claim 4 . wherein the direction of change in the disease state includes at least one of different colors, different densities, different numbers, and different symbols for each pixel unit or each pixel group unit; generating the disease state change map so that it is identifiably delineated;
The medical image processing apparatus according to claim 6 . a storage unit that stores a lookup table in which the type of the texture pattern and the degree of progression of the disease state are associated;
The estimation unit estimates the direction of change in the disease state of the subject by referring to the lookup table.
The medical image processing apparatus according to any one of claims 4 to 7 . an acquisition unit that acquires a plurality of images obtained by imaging the same subject at different imaging times;
By analyzing the tissue properties of the subject based on the pixel values of the respective regions of the plurality of images, the tissue properties of the subject are classified into a plurality of tissue property classes, and the classified tissue property classes are classified into , an analysis unit that allocates to each region of the plurality of images;
an estimating unit that estimates changes in the disease state of the subject from changes in the tissue attribute class in corresponding regions of the plurality of images,
The analysis unit performs texture analysis based on pixel values of each region of the plurality of images, classifies each region into one of a plurality of texture patterns, and classifies the classified texture patterns as the assigned to each region of a plurality of images,
The estimating unit estimates a change in the disease state in the corresponding region and a change speed of the disease state from the change in the texture pattern in the corresponding region of each of the plurality of images,
The estimating unit further refers to a lookup table in which the type of texture pattern and the degree of progress of the disease state are associated to determine the speed of recovery in the direction of recovery and the speed of exacerbation in the direction of exacerbation. further estimate at least one
Medical image processing equipment. an acquisition unit that acquires a plurality of images obtained by imaging the same subject at different imaging times;
By performing texture analysis based on pixel values of respective regions of the plurality of images, each region is classified into one of a plurality of texture patterns, and the classified texture patterns are assigned to each of the plurality of images. an analysis unit that allocates to the region of
an estimating unit for estimating changes in the disease state of the subject from changes in tissue attribute classes in corresponding regions of each of the plurality of images;
A memory for storing a plurality of lookup tables that are disease cause-specific lookup tables respectively corresponding to different disease causes and in which the type of the texture pattern and the degree of progression of the disease state are associated for each disease cause. and
The estimating unit estimates the disease state of the subject by referring to the lookup table for each disease cause corresponding to the cause of disease specified by the user from among the plurality of lookup tables for each disease cause. estimating the direction of change,
Medical image processing equipment. an acquisition unit that acquires a plurality of images obtained by imaging the same subject at different imaging times;
By performing texture analysis based on pixel values of respective regions of the plurality of images, each region is classified into one of a plurality of texture patterns, and the classified texture patterns are assigned to each of the plurality of images. an analysis unit that allocates to the region of
an estimating unit for estimating changes in the disease state of the subject from changes in tissue attribute classes in corresponding regions of each of the plurality of images;
A memory for storing a plurality of lookup tables that are disease cause-specific lookup tables respectively corresponding to different disease causes and in which the type of the texture pattern and the degree of progression of the disease state are associated for each disease cause. and
The estimating unit estimates the disease state of the subject by referring to the lookup table for each disease cause corresponding to the cause of disease specified by the user from among the plurality of lookup tables for each disease cause. Estimate the direction of change,
The estimating unit further estimates the cause of the disease of the subject from changes in the types of the texture patterns in the local regions corresponding to each other in the plurality of images .
Medical image processing equipment. the lookup table is configured to allow its contents to be created, edited or changed by a user;
The medical image processing apparatus according to any one of claims 8 to 11 . an acquisition unit that acquires a plurality of images obtained by imaging the same subject at different imaging times;
By analyzing the tissue properties of the subject based on the pixel values of the respective regions of the plurality of images, the tissue properties of the subject are classified into a plurality of tissue property classes, and the classified tissue property classes are classified into , an analysis unit that allocates to each region of the plurality of images;
an estimating unit that estimates changes in the disease state of the subject from changes in the tissue attribute class in corresponding regions of the plurality of images ,
The analysis unit performs texture analysis based on pixel values of each region of the plurality of images, classifies each region into one of a plurality of texture patterns, and classifies the classified texture patterns as the assigned to each region of a plurality of images,
The estimating unit estimates the change in the disease state in the corresponding region and the rate of change in the disease state from the change in the texture pattern in the corresponding region of each of the plurality of images.
Medical image processing system. A control method for a medical image processing apparatus comprising a processing circuit, comprising:
The processing circuit is
Acquiring a plurality of images obtained by imaging the same subject at different imaging times,
classifying the tissue properties of the subject into a plurality of tissue property classes by analyzing the tissue properties of the subject based on the pixel values of the respective regions of the plurality of images;
assigning the classified tissue attribute classes to respective regions of the plurality of images;
estimating a change in the disease state of the subject from a change in the tissue attribute class in the corresponding region of each of the plurality of images ;
By performing texture analysis based on pixel values of respective regions of the plurality of images, each region is classified into one of a plurality of texture patterns, and the classified texture patterns are assigned to each of the plurality of images. allocated to the area of
estimating the change in the disease state in the corresponding region and the rate of change in the disease state from the change in the texture pattern in the corresponding region of each of the plurality of images;
controlling the processing circuitry to
A control method for a medical image processing apparatus .

Images