Image processing device, image processing method and program

Abstract

To enable a user such as a physician to perform image reading using differences without burden and without being affected by noise.SOLUTION: An image processing device according to an embodiment includes: an image acquisition unit that acquires a first image including a first region image obtained by imaging a common subject and a second image different from the first image and including a second region image corresponding to the first region image; a first reliability estimation unit that estimates the reliability of a first pixel value on the basis of the first pixel value constituting the first region image and a first reference value; a second reliability estimation unit that estimates the reliability of a second pixel value on the basis of the second pixel value constituting the second region image and a second reference value different from the first reference value; and a difference value reliability calculation unit that acquires the reliability of the difference value between the first pixel value and the second pixel value on the basis of the reliability of the first pixel value and the reliability of the second pixel value.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] Embodiments of the present invention relate to an image processing apparatus, an image processing method, and a program.

Background Art

[0002] In the medical field, diagnoses are made using images obtained by various medical imaging devices (modalities) such as computed tomography (CT) devices. For efficient diagnosis, a technique is needed to visualize the temporal changes of lesions to assist doctors in diagnosis.

[0003] For example, in order to grasp the risk of bone-related events in bone metastasis, a technique is needed to visualize the intraspinal canal invasion of a tumor and measure the degree of invasion. As a technique for visualizing temporal changes, an image difference technique is known that supports comparison between images by aligning two images taken at different times and displaying a difference image that visualizes the differences between the images.

[0004] As noise that easily occurs in a region surrounded by a structure having a high absorption coefficient such as bone, beam hardening artifacts are known. Beam hardening occurs when a polychromatic beam passes through a patient, and soft (low energy) photons are preferentially absorbed or scattered from the beam, leaving hard photons. This beam hardening causes a difference in the linear attenuation coefficient because the optical density of the optical path passing through the patient differs depending on the position of the X-ray source and the exposure direction of the X-ray during scanning when the X-ray source and the detector are synchronously rotated and scanned in the gantry of the CT device to image the patient. That is, the beam hardening artifact occurs when the CT value in the direction passing through the high absorber becomes weak and a region with a low CT value appears during reconstruction.

[0005] Based on the correlation between two images obtained by imaging with X-rays of different tube voltages, for each pixel in the artifact region, the pixel value in the high-tube-voltage image when there is no artifact is estimated to remove noise, and the technique is disclosed in Patent Document 1.

Prior Art Documents

Patent Documents

[0006]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0007] However, in the technique of Patent Document 1, in addition to the need for two imaging at different tube voltages, due to the influence of the image characteristics of the high-tube-voltage image, there is a risk that the image features specific to the lesion will decrease. For this reason, there is a problem that the lesion location cannot be appropriately depicted in the difference image.

[0008] One of the problems to be solved by the embodiments disclosed in this specification and the drawings is to allow users such as doctors to perform reading using differences without load by excluding the influence of noise. However, the problems to be solved by the embodiments disclosed in this specification and the drawings are not limited to the above problems. The problems corresponding to the respective effects of the respective configurations shown in the embodiments described later can also be positioned as other problems.

Means for Solving the Problems

[0009] The image processing apparatus according to the embodiment includes an image acquisition unit that acquires a first image including a first region image in which a common subject is imaged, and a second image different from the first image including a second region image corresponding to the first region image, a first reliability estimation unit that estimates the reliability of the first pixel value based on the first pixel value constituting the first region image and a first reference value, a second reliability estimation unit that estimates the reliability of the second pixel value based on the second pixel value constituting the second region image and a second reference value different from the first reference value, and a difference value reliability calculation unit that acquires the reliability of the difference value between the first pixel value and the second pixel value based on the reliability of the first pixel value and the reliability of the second pixel value.

Brief Description of the Drawings

[0010]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14A

Figure 14B

Figure 15

Figure 16

Figure 17

Figure 18

[0011] Hereinafter, embodiments of an image processing apparatus, an image processing method, and a program will be described in detail with reference to the accompanying drawings.

[0012] Hereinafter, the present invention will be described in detail based on its preferred embodiments with reference to the accompanying drawings. Note that the configurations shown in the following embodiments are merely examples, and the present invention is not limited to the illustrated configurations.

[0013] <First Embodiment> In the first embodiment, as an example, details of an image processing apparatus for assisting in the diagnosis of cancer bone metastasis-related events will be described. In this diagnostic assistance, bone metastasis, intraspinal canal invasion, and extraosseous tumors around vertebrae are the diagnostic targets, and a differential image (differential data) appropriate for visualizing these in the same image is generated. In this description, the spinal cord region within the spinal canal is referred to as the spinal canal region.

[0014] The image processing apparatus according to this embodiment is an apparatus for visualizing a candidate region for intraspinal canal invasion and a candidate region for extraosseous tumor in a subject region shown in a target image to be processed. Specifically, first, the image processing apparatus generates a differential image between two time-series images that are target images. Next, the image processing apparatus extracts the spinal canal from the target image, and further extracts an infiltration candidate region as a candidate region for intraspinal canal invasion, which is an abnormal region within the spinal canal, using the differential image. Also, the image processing apparatus extracts the region around the vertebra from the target image, and further extracts a tumor candidate region as a candidate region for extraosseous tumor, which is an abnormal region within the region around the vertebra, using the differential image. Then, the image processing apparatus performs control to calculate and display as differential data a differential image, an image in which the infiltration candidate region and the tumor candidate region are superimposed on the target image, and a graph of index values representing the degree of infiltration and tumor.

[0015] FIG. 1 is a diagram showing an example of the configuration of an image processing apparatus 10 according to the first embodiment. As shown in FIG. 1, the image processing apparatus 10 includes a processing circuit 11, a communication interface 12, a storage circuit 13, a display 14, an input interface 15, and a connection unit 16. Also, the image processing apparatus 10 is communicably connected via a network (not shown) to a medical image diagnostic apparatus (modality), a medical image storage apparatus, each department system, and the like.

[0016] Medical imaging diagnostic devices include X-ray CT (Computed Tomography) devices, magnetic resonance imaging (MRI) devices, X-ray diagnostic devices, etc. The medical image storage device is realized by a PACS (Picture Archiving and Communication System) or the like and stores medical images in a format compliant with DICOM (Digital Imaging and Communications in Medicine). Each department system includes various systems such as a hospital information system (HIS), a radiology information system (RIS), a diagnostic report system, and a clinical laboratory information system (LIS).

[0017] The processing circuit 11 controls the image processing apparatus 10 by executing a control function 11a, an image acquisition function 11b, a first reliability estimation function 11c, a second reliability estimation function 11d, a difference value reliability calculation function 11e, and a difference data calculation function 11f in response to an input operation received from a user via the input interface 15. Further, the processing circuit 11 executes an alignment function 11b1 and a region extraction function 11b2. Here, the image acquisition function 11b is an example of an image acquisition unit. Also, the first reliability estimation function 11c is an example of a first reliability estimation unit. Also, the second reliability estimation function 11d is an example of a second reliability estimation unit. Also, the difference value reliability calculation function 11e is an example of a difference value reliability calculation unit. Also, the difference data calculation function 11f is an example of a difference data calculation unit.

[0018] The control function 11a controls to generate various GUIs (Graphical User Interfaces) and various display information according to operations via the input interface 15 and display them on the display 14. Also, the control function 11a controls the transmission and reception of information with devices and systems on a network (not shown) via the communication interface 12. Further, the control function 11a acquires information about the subject from each department system connected to the network. Moreover, the control function 11a outputs the processing result to devices and systems on the network.

[0019] The image acquisition function 11b acquires medical image data obtained by photographing a subject with a medical imaging diagnostic device or the like from a medical image storage device or the like based on image reference address information (URL (Uniform Resource Locator), various UIDs (Unique Identifiers), Path strings, etc.). Specifically, the image acquisition function 11b acquires three-dimensional medical images (volume data) from the modality connected to the image processing device 10 and the network, the medical image storage device, etc. Note that the processing by the image acquisition function 11b will be described in detail later. The medical image is an example of an image. In the following description, the medical image may be simply referred to as an image.

[0020] The alignment function 11b1 performs alignment processing on the three-dimensional image acquired by the image acquisition function 11b. Specifically, the alignment function 11b1 calculates the spatial correspondence between the first image and the second image. Then, it generates a deformed image obtained by coordinate transformation by aligning the second image with the first image. Note that the processing by the alignment function 11b1 will be described in detail later.

[0021] The region extraction function 11b2 extracts the spinal canal region from the image, targeting the three-dimensional image acquired by the image acquisition function 11b or the deformed image generated by the alignment function 11b1. The processing by the region extraction function 11b2 will be described in detail later.

[0022] The first reliability estimation function 11c estimates, using a predetermined reference value, the reliability indicating the degree of regarding each pixel value of the first image acquired by the image acquisition function 11b as a CT value representing a lesion due to infiltration. The processing by the first reliability estimation function 11c will be described in detail later.

[0023] The second reliability estimation function 11d estimates, using a reference value different from that of the first reliability estimation function 11c, the reliability indicating the degree of regarding each pixel value of the second image acquired by the image acquisition function 11b or each pixel value of the deformed image generated by the alignment function 11b1 as the true CT value not affected by noise. The processing by the second reliability estimation function 11d will be described in detail later.

[0024] The difference value reliability calculation function 11e calculates the reliability of the difference value between the first pixel value and the second pixel value using the reliability of the first pixel value estimated by the first reliability estimation function 11c and the reliability of the second pixel value estimated by the second reliability estimation function 11d. The processing by the difference value reliability calculation function 11e will be described in detail later.

[0025] The difference data calculation function 11f calculates the difference data between the first pixel value and the second pixel value. The processing by the difference data calculation function 11f will be described in detail later.

[0026] The above-described processing circuit 11 is realized by, for example, a processor. In that case, each of the above-described processing functions is stored in the storage circuit 13 in the form of a program executable by a computer. Then, the processing circuit 11 reads and executes each program stored in the storage circuit 13 to realize the functions corresponding to the respective programs. In other words, the processing circuit 11 has each of the processing functions shown in FIG. 1 when each program is read.

[0027] Note that the processing circuit 11 may be configured by combining a plurality of independent processors, and each processing function may be realized by each processor executing a program. Further, each processing function included in the processing circuit 11 may be appropriately distributed or integrated into a single or a plurality of processing circuits and realized. Further, each processing function included in the processing circuit 11 may be realized by a combination of hardware such as a circuit and software. Here, an example in which a program corresponding to each processing function is stored in a single storage circuit 13 has been described, but the embodiment is not limited to this. For example, the programs corresponding to each processing function may be distributed and stored in a plurality of storage circuits, and the processing circuit 11 may be configured to read and execute each program from each storage circuit.

[0028] The communication interface 12 controls the transmission and communication of various data transmitted and received between the image processing apparatus 10 and other apparatuses and systems connected via a network. Specifically, the communication interface 12 is connected to the processing circuit 11, and outputs data received from other apparatuses and systems to the processing circuit 11, or transmits data output from the processing circuit 11 to other apparatuses and systems. For example, the communication interface 12 is realized by a network card, a network adapter, a NIC (Network Interface Controller), or the like.

[0029] The storage circuit 13 stores various data and various programs. Specifically, the storage circuit 13 is connected to the processing circuit 11, stores data input from the processing circuit 11, or reads out stored data and outputs it to the processing circuit 11. For example, the storage circuit 13 is realized by a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory, a hard disk, an optical disk, or the like.

[0030] The display 14 displays various types of information and various data. Specifically, the display 14 is connected to the processing circuit 11 and displays various types of information and various data output from the processing circuit 11. For example, the display 14 is realized by a liquid crystal display, a CRT (Cathode Ray Tube) display, an organic EL display, a plasma display, a touch panel, or the like. The display 14 is an example of a display unit.

[0031] The input interface 15 receives input operations of various instructions and various information from the user. Specifically, the input interface 15 is connected to the processing circuit 11, converts the input operation received from the user into an electrical signal, and outputs it to the processing circuit 11. For example, the input interface 15 is realized by a trackball, a switch button, a mouse, a keyboard, a touch pad that performs an input operation by touching an operation surface, a touch screen in which a display screen and a touch pad are integrated, a non-contact input interface using an optical sensor, a voice input interface, or the like. Note that in this specification, the input interface 15 is not limited to only those having physical operation components such as a mouse and a keyboard. For example, an electrical signal processing circuit that receives an electrical signal corresponding to an input operation from an external input device provided separately from the apparatus and outputs this electrical signal to the control circuit is also included in the example of the input interface 15.

[0032] The connection unit 16 is a bus or the like that connects the processing circuit 11, the communication interface 12, the storage circuit 13, the display 14, and the input interface 15.

[0033] In this specification, an axis representing the direction from the right hand to the left hand of the subject is defined as the X-axis, an axis representing the direction from the front to the back of the subject is defined as the Y-axis, and an axis representing the direction from the head to the feet of the subject is defined as the Z-axis. Also, the XY cross-section is defined as the axial plane, the YZ cross-section is defined as the sagittal plane, and the ZX cross-section is defined as the coronal plane. That is, the X-axis direction is the direction orthogonal to the sagittal plane (hereinafter referred to as the sagittal direction). Also, the Y-axis direction is the direction orthogonal to the coronal plane (hereinafter referred to as the coronal direction). Furthermore, the Z-axis direction is the direction orthogonal to the axial plane (hereinafter referred to as the axial direction). At this time, in the case of a CT image configured as a set of two-dimensional tomographic images (slice images), the slice plane of the image represents the axial plane, and the direction orthogonal to the slice plane (hereinafter referred to as the slice direction) represents the axial direction. Note that the way of taking the coordinate system is an example, and other definitions may be used.

[0034] FIG. 2 is a flowchart showing the processing steps executed by the image processing apparatus 10 according to the first embodiment. Also, FIG. 3 is a data flow diagram showing the flow of all data of the image processing apparatus 10 according to the first embodiment.

[0035] The image processing apparatus 10 inputs a current image that is a first image and a past image that is a second image, and outputs a difference image showing the difference between the current image and the past image as difference data. At this time, the reliability of each pixel of the input images is calculated, and based on this, noise suppression processing is performed on the difference image. Here, all the images are three-dimensional volume data. In this embodiment, the first image and the second image will be described as X-ray CT images. However, the embodiment is not limited to this, and the present invention is applicable to medical images of other modalities such as MRI (Magnetic Resonance Imaging) images and ultrasonic images. Also, the first image and the second image may be a combination other than the current image and the past image. For example, they may be a contrast image and a non-contrast image taken in the same day's examination.

[0036] The process of step S2010 shown in FIG. 2 is realized, for example, by the processing circuit 11 calling and executing a program corresponding to the image acquisition function 11b from the storage circuit 13. Also, the process of step S2011 is realized, for example, by the processing circuit 11 calling and executing a program corresponding to the alignment function 11b1 from the storage circuit 13. Also, the processes of steps S2012 and S2013 are realized, for example, by the processing circuit 11 calling and executing a program corresponding to the region extraction function 11b2 from the storage circuit 13. Also, the process of step S2020 is realized, for example, by the processing circuit 11 calling and executing a program corresponding to the first reliability estimation function 11c from the storage circuit 13. Further, the process of step S2030 is realized, for example, by the processing circuit 11 calling and executing a program corresponding to the second reliability estimation function 11d from the storage circuit 13. Further, the process of step S2040 is realized, for example, by the processing circuit 11 calling and executing a program corresponding to the difference value reliability calculation function 11e from the storage circuit 13. Further, the process of step S2050 is realized, for example, by the processing circuit 11 calling and executing a program corresponding to the difference data calculation function 11f from the storage circuit 13.

[0037] (Step S2010) As shown in FIG. 3, in the image acquisition unit 201, the current image, which is the first image, is acquired from the PACS or the modality using the image acquisition function 11b, the past image of the same patient is identified using the method described later, and the past image, which is the second image, is acquired from the PACS.

[0038] Here, the images to be acquired are described below as three-dimensional volume data obtained by photographing the same part of a certain patient, but this acquisition function is not limited thereto, and for example, two-dimensional pixel data may be used.

[0039] As a method for identifying a past image of the same patient, for example, there is a method of searching from PACS or the like for an image that has the same patient ID as the first image and whose examination date is older than that of the first image, and selecting the image with the most recent examination date from among them. This acquisition function may identify past images by other methods or by manual selection by the user.

[0040] In this embodiment, the second image will be described by taking as an example the case where it is an image taken in the past compared to the first image. However, the implementation of the present invention is not limited to this. For example, the first image may be an image taken in the past compared to the second image, or two images taken at the same time may be used as the first image and the second image. The first image and the second image may be images with different contrast phases. Also, the first image and the second image may be images of different patients.

[0041] (Step S2011) Furthermore, the image acquisition unit 201 performs alignment between the current image and the past image using the alignment function 11b1, and calculates the spatial correspondence relationship between the current image and the past image.

[0042] Here, the alignment function 11b1 can calculate (estimate) the spatial correspondence relationship between images using an existing linear alignment algorithm, a non-linear alignment algorithm, or a combination of them. The alignment function 11b1 can align the positions of the feature points indicating the characteristic parts included in the current image and the feature points indicating the characteristic parts included in the past image by the above-described deformed alignment between images.

[0043] Furthermore, based on the result of the alignment, the alignment function 11b1 generates a deformed image obtained by coordinate-transforming the past image to match the current image.

[0044] In the above example, the case where alignment processing is performed between the current image and the past image, and a deformed image is generated based on the result was described as an example. However, the implementation of the present invention is not limited to this. For example, a deformed image generated by coordinate-transforming a past image by performing the above-described processing may be stored in a PACS or the like, and the image acquisition unit 201 may acquire the image. Further, an image obtained by photographing after adjusting the position and posture of the subject in advance so as to match the past image may be acquired as the current image. In this case, since the current image and the past image have a spatial correspondence relationship, generation of a deformed image becomes unnecessary. That is, in the subsequent processing, the past image can be used as it is instead of the deformed image.

[0045] (Step S2012) Furthermore, in the image acquisition unit 201, segmentation of the spinal canal is performed on the current image using the region extraction function 11b2, and first target region information for specifying a first target region, that is, spinal canal region information, is calculated.

[0046] Here, the region extraction function 11b2 performs segmentation of the spinal canal by a known region extraction method. For example, segmentation is performed by a threshold process, a morphological operation, pattern matching, an inference device such as U-Net, and the spinal canal region information of the current image is extracted. However, the region extraction function 11b2 may acquire the spinal canal region information in the current image by input by the user.

[0047] (Step S2013) Furthermore, in the image acquisition unit 201, segmentation of the spinal canal is also performed on the deformed image using the region extraction function 11b2, and second target region information for specifying a second target region, that is, spinal canal region information of the deformed image, is calculated.

[0048] Here, in addition to performing segmentation of the spinal canal by a known region extraction method, the region extraction function 11b2 may use the correspondence relationship between the pixels of the current image and the past image calculated by the alignment function 11b1 to convert the spinal canal region information of the current image and calculate the spinal canal region information of the deformed image.

[0049] In FIG. 2, step S2020 and step S2030 are executed as sequential processes, but they may be executed as parallel processes.

[0050] (Step S2020) As shown in FIG. 3, in the first reliability estimation unit 202, based on the current image which is the first image acquired by the image acquisition unit 201 and the spinal canal region information of the first image, the reliability of the first pixel value representing the degree to which each pixel (first pixel) in the spinal canal region of the first image is a pixel within the lesion region is estimated. Specifically, for each of a plurality of pixels (first pixels) in the spinal canal region of the first image, using the first pixel value a1 which is the pixel value of the pixel and the standard pixel value m1, the reliability b1 of the first pixel value is calculated by the following formula (1). Here, as the standard pixel value m1, the average pixel value of the spinal canal region of the first image may be used, or a representative value (about 40 HU) as the average CT value in the spinal canal of a healthy person may be used. Also, a preset value other than the average pixel value of the spinal canal region may be used.

[0051]

Equation

[0052] Here, W(a, μ) is an activation function that smoothly decays when a is small with a reference value μ as shown in, for example, FIG. 10. This function is also used when calculating the reliability of the second pixel value in step S2030 described later, but different reference values are used for the calculations. This is because the reliability of the first pixel value calculates the reliability indicating that the pixel is a lesion, while the reliability of the second pixel value calculates the reliability indicating that the pixel is not affected by noise. The first reliability estimation unit 202 calculates the reliability of the first pixel value by utilizing the characteristic that the pixel value of the lesion due to infiltration tends to be higher than the average pixel value of the pixels including the surrounding area other than the lesion. Therefore, the reference value μ1 used for calculating the reliability of the first pixel value is set to be larger (for example, about 40 HU) than the reference value μ2 used for calculating the second reliability described later. Specifically, a value corresponding to the difference between the pixel value of the lesion due to infiltration and the average pixel value in the spinal canal is set.

[0053] The above method for calculating the reliability of the first pixel value is only an example, and the first reliability estimation unit 202 may calculate the reliability of the first pixel value by other methods. For example, as an activation function, the first reliability estimation unit 202 may use a function in which the function W linearly decreases as a decreases with the reference value μ as a boundary as shown in the following formula (2), and the value range is from 0 to 1. FIG. 18 shows a graph of the following activation function W.

[0054]

Equation

[0055] In the above example, a case was described in which lesions presenting CT values higher than the average CT value in a healthy spinal canal were considered, and the reliability of the region where such lesions exist was calculated with high accuracy. However, the embodiments are not limited to this. For example, the first reliability estimation unit 202 may consider lesions presenting CT values lower than the average CT value in a healthy spinal canal, and calculate the reliability of the region where such lesions exist with high accuracy. Further, the first reliability estimation unit 202 may calculate the reliability of the first pixel value based on the statistical distribution of the CT values in a healthy spinal canal and based on the degree of deviation from the distribution.

[0056] (Step S2030) As shown in FIG. 3, in the second reliability estimation unit 203, based on the deformed image obtained by coordinate transformation of the past image, which is the second image acquired by the image acquisition unit 201, and the spinal canal region information of the deformed image, the reliability of the second pixel value representing the degree to which each pixel (second pixel) in the spinal canal region of the deformed image is a pixel outside the noise region is estimated. Specifically, for each of a plurality of pixels (second pixels) in the spinal canal region of the deformed image, the reliability b2 of the second pixel value is calculated by the following formula (3) using the second pixel value a2, which is the pixel value of the pixel, and the standard pixel value m2. Here, as the standard pixel value m2, the average pixel value of the spinal canal region of the second image may be used, or a representative value (about 40 HU), which is the average CT value in the spinal canal when not affected by beam hardening artifacts, may be used. Further, a preset value other than the average pixel value of the spinal canal region may be used.

[0057]

Equation

[0058] Here, W(a, μ) is an activation function that smoothly decays when a is small with respect to a reference value μ as shown in, for example, FIG. 10. Pixel values in the area affected by beam hardening artifacts tend to be smaller than those in the area not affected by the influence. Therefore, in order to reduce the reliability of the pixel values at the sites where beam hardening artifacts occur, the reference value μ2 is set to a smaller value (for example, about 0 HU) than in the case of the first reliability estimation unit 202.

[0059] The above method for calculating the reliability of the second pixel value is merely an example, and the second reliability estimation unit 203 may calculate the reliability of the second pixel value by other methods. For example, the second reliability estimation unit 203 may use, as an activation function, a function in which a linearly decreases with respect to a reference value μ as shown in the following formula (4).

[0060]

Equation

[0061] In the above example, as a factor of noise that the second pixel may be affected by, the method in the case of considering beam hardening artifacts has been described, but the embodiments are not limited thereto. For example, the second reliability estimation unit 203 may calculate the reliability of the second pixel value to be low when the pixel value is higher than the average CT value in the spinal canal in consideration of the high luminance of the pixel due to metal artifacts. In addition to the factors of noise exemplified above, the second reliability estimation unit 203 may calculate the reliability of the second pixel value in consideration of the effects of motion artifacts, ring artifacts, banding artifacts, etc. Further, the second reliability estimation unit 203 is not limited to the method of considering a single noise factor, and may calculate the reliability of the second pixel value in consideration of a plurality of noise factors at the same time. Specifically, the second reliability estimation unit 203 may calculate the reliability independently for a plurality of noise factors and integrate (for example, perform a product operation) the calculated values to calculate the reliability of the second pixel value. The second reliability estimation unit 203 may also calculate the reliability of the second pixel value based on the deviation degree from the statistical distribution of the CT values in the spinal canal when not affected by noise.

[0062] (Step S2040) In the difference value reliability calculation unit 204, using the reliability b1 of the first pixel value and the reliability b2 of the second pixel value, the reliability b of the difference value between two images calculated in step S2050 described later d is calculated. Here, the pixel value of the first pixel is the first pixel value, and the pixel value of the second pixel is the second pixel value. Also, the first pixel and the second pixel are a pixel pair associated by the alignment function 11b1 in step S2011. The reliability b d is specifically calculated by the following formula (5).

[0063]

Equation

[0064] In the above formula (5), if either one of the reliability b1 and the reliability b2 is large, the reliability b of the difference valued It also becomes larger and is configured so that its maximum value is suppressed to 1 or less. If it is calculated based on the reliability b1 of the first pixel value and the reliability b2 of the second pixel value, the calculation formula for the reliability of the difference value is not limited to this. For example, it can also be configured with an expression using the square root sqrt() as in the following formula (6).

[0065]

Equation

[0066] The difference value output from the image processing apparatus 10 is the result of multiplying the original difference value by the reliability b of this difference value. d By doing so, as shown in FIG. 17, it is possible to increase the reliability of the difference value in the range 171 where infiltration exists and decrease the reliability of the difference value in the range 173 where noise exists. That is, when either the reliability of the first pixel value or the reliability of the second pixel value is high, the reliability of the difference value is increased. This means increasing the reliability of the difference value when the pixel of the current image is likely to be a lesion or when the pixel of the past image is likely not to be affected by noise. On the other hand, when both of these reliabilities are low, the reliability of the difference value is decreased. That is, it means decreasing the reliability of the difference value when the pixel value of the current image is unlikely to be a lesion and the pixel of the past image is likely to be affected by noise. Here, the reason why the range 172 for attenuating the difference is high (towards the right) in the pixel value of the current image is to increase the reliability of the difference value when there is infiltration in the current image even if there is noise in the past image. In this processing step, based on the reliability of the first pixel value and the reliability of the second pixel value, the reliability of the difference value is calculated with the above intention.

[0067] In the differential value reliability calculation unit 204, the differential value reliability may be further corrected based on the position of each pixel in the patient's body. For example, correction of the differential value reliability according to the intervertebral foramen level (correction 1) is performed by the following formula (7), and the reliability b of the differential value is such that the differential value of a clinically important site is left as it is. d ´ may be calculated.

[0068]

Equation

[0069] Here, C(z) is a weighting coefficient for correction that takes values as shown in, for example, FIG. 11 according to the intervertebral foramen level to which the patient coordinate z belongs. Here, the intervertebral foramen level represents the division of the z - coordinate position centered on the intervertebral foramen in the cranio - caudal direction along the spine, and is labeled with intervertebral foramen labels such as C1, C2, …, C8, T1, T2, …, T12, L1, L2, …, L5. The value of C(z) can be obtained, for example, by the differential value reliability calculation function 11e reading the values stored in a table for each intervertebral foramen label from the storage circuit 13 and performing interpolation if necessary. By this correction, in clinically important sites caudal to T3, the contribution rate of reliability is set low, and the raw differential value without noise reduction processing can be directly seen by the doctor. This makes it possible to prevent unintentional processing associated with noise reduction processing from entering the differential values of clinically important sites.

[0070] In the differential value reliability calculation unit 204, further correction of the differential value reliability according to the intervertebral foramen level (correction 2) is performed by the following formula (8), and the reliability b of the differential value is such that the differential values of sites that are not very clinically important and are prone to noise due to beam hardening artifacts are attenuated. d ´´ may be calculated.

[0071]

Equation

[0072] Here, C´(z) is a correction weight coefficient that takes values as shown in, for example, FIG. 12 according to the intervertebral foramen level to which the patient coordinate z belongs.

[0073] As described above, the correction of the difference value reliability according to the intervertebral foramen level has been explained. However, the weighted value may be anything as long as it is represented corresponding to the patient coordinate z. For example, the correction of the difference value reliability according to the vertebral level can be similarly performed. Here, the vertebral level represents the division of the z - coordinate position centered on the vertebra in the cranio - caudal direction along the spine, and vertebral labels such as C1, C2, …, C7, T1, T2, …, T12, L1, L2, …, L5 are attached. Similar to the case of the intervertebral foramen label, it can be implemented by referring to the table stored in the memory circuit 13 by the difference value reliability calculation function 11e.

[0074] In addition to the method of performing correction by referring to the table stored in the memory circuit 13, for example, the difference value reliability calculation function 11e can also directly calculate the weight coefficient required for correction 1 from the z - coordinate value by the following formula (9).

[0075]

Equation

[0076] Here, Z(T3) is the z - coordinate value of the intervertebral foramen T3, and Z(T2) is the z - coordinate value of the intervertebral foramen T2. The difference value reliability calculation function 11e uses this weight coefficient C L (z) to calculate the reliability b d ’ of the corrected difference value by the following formula (10).

[0077]

Equation

[0078] In the above example, the case where the differential reliability is corrected based on the position in the patient's body was described as an example. However, the method of correcting the reliability is not limited to this. For example, the differential value reliability calculation function 11e may correct the values of the first reliability and the second reliability based on the position in the patient's body. Specifically, the differential value reliability calculation function 11e may correct the first reliability to be higher than other parts at the position from the lower thoracic vertebrae to the upper lumbar vertebrae where infiltration is likely to occur in the spinal canal. Also, the differential value reliability calculation function 11e may correct the second reliability to be lower than other parts at the position from the upper thoracic vertebrae to the cervical vertebrae where the occurrence frequency of beam hardening is high.

[0079] In addition, the correction of the reliability based on the position in the patient's body is not limited to the method based on the position in the cranio-caudal direction of the patient as described above, and may be a method based on the position on the dorsal side or ventral side of the patient, or may be a method based on the position on the left side or right side of the patient.

[0080] Note that the reliability correction process as described above does not necessarily have to be performed.

[0081] Furthermore, the differential value reliability calculation function 11e may calculate the differential value reliability according to the vertebral level or the intervertebral foramen level without using the reliability b1 of the first pixel value and the reliability b2 of the second pixel value. In this case, the differential value reliability calculation function 11e sets b d = 1 and performs the above correction 1 and correction 2.

[0082] (Step S2050) The differential data calculation unit 205 calculates a differential image based on the first image, the deformed image, and the reliability of the differential value. At this time, the differential data calculation unit 205 uses the first pixel value a1 of the first image, the second pixel value a2 of the deformed image, and the reliability b of the differential value between the first pixel value and the second pixel value dTherefore, for example, the pixel value d of the difference image is calculated using the following formula (11). Here, the first pixel presenting the first pixel value a1 and the second pixel presenting the second pixel value a2 are a pixel pair associated in step S2011. Accordingly, the difference data calculation unit 205 can attenuate the difference value in a region with low reliability based on the reliability calculated in step S2040.

[0083] [Number]

[0084] Note that the method for calculating the pixel value d is not limited to the above example. The difference data calculation unit 205 may perform a non-linear transformation on the reliability b of the difference value. d For example, when the reliability b of the difference value is higher than a predetermined reference value, the difference data calculation unit 205 may calculate (a1 - a2) as the pixel value, and when it is not, may calculate 0 as the pixel value. Also, the difference data calculation unit 205 may use anything as long as it is "a value representing the difference in pixel values" instead of the simple difference (a1 - a2). For example, the difference data calculation unit 205 may calculate a value representing the difference in pixel values based on the pixel values of the peripheral pixels of the first pixel and the second pixel by the Voxel Matching method or the like. d

[0085] The difference data calculation unit 205 may further assign a specified color to the magnitude of the pixel value d of the difference image, and calculate a fused image obtained by superimposing the assigned color on the current image as the difference data.

[0086] The difference data calculation unit 205 may further calculate the occupancy rate g(z) from the first target region I(z) within the slice including the patient coordinate z and the threshold T for determining infiltration using the spinal canal region information which is the first target region information, according to the following formula (12). Here, [(proposition)] is the Iverson notation, which returns 1 when (proposition) is true and 0 when false.

[0087] [Number]

[0088] The occupancy rate calculated in this way is calculated for each of the patient coordinates z by the differential data calculation unit 205, and an image obtained by arranging the resulting graph alongside the sagittal cross-section or projection image of the fusion image as shown in FIG. 13 may be generated as occupancy rate data and displayed.

[0089] In FIG. 13, I sag represents the image of the sagittal plane in the target image. In the image I sag , R ver is the vertebral region, R spi is the spinal canal region, R inf is the region of intraspinal canal invasion (intraspinal canal invasion region). G idx represents the occupancy rate graph. On each axial plane of the target image, the occupancy rate g(z) representing the degree of invasion into the spinal canal is plotted along the Z-axis of the image I sag . In P1, the occupancy rate graph is high due to intraspinal canal invasion, and in P2, the occupancy rate graph is high due to noise caused by beam hardening artifacts. G' idx represents the occupancy rate graph when the reliability of the difference value is not used, and the occupancy rate graph in P2 is higher than that of G idx .

[0090] The differential data calculation unit 205 outputs the differential image, fusion image, and occupancy rate data calculated as described above as differential data to an external PACS or the like.

[0091] As described above, according to the image processing apparatus 10 according to the first embodiment, it is possible to cause a user such as a doctor to perform reading using the difference without being affected by noise and without a load. That is, by using the image processing apparatus 10 according to the first embodiment, it is possible to solve the problem that reading using the difference value cannot be appropriately performed due to the influence of noise.

[0092] (Modification Example 1: High-precision spinal canal segmentation at the time of image acquisition) For the first embodiment, in the image acquisition unit 201, the spinal canal region information protruding into the gap between the nerve roots as shown in FIG. 14A or the spinous processes as shown in FIG. 14B may be removed. Specifically, regarding the nerve roots as shown in FIG. 14A, the image acquisition unit 201 detects the intervertebral foramen, which is the intersection hole of the spinal canal, increases the size of the structural element for the opening operation as it approaches the intervertebral foramen, and can achieve highly accurate segmentation by strongly removing the protruding parts from the spinal canal region information.

[0093] Also, regarding the gap between the spinous processes as shown in FIG. 14B, the image acquisition unit 201 can achieve highly accurate segmentation by removing the locations with low CT values while changing the threshold according to the location. Specifically, in the cervical vertebrae and upper thoracic vertebrae, since the gap between the spinous processes is not large, the image acquisition unit 201 lowers the threshold value of the CT value to be excluded and performs weak removal.

[0094] Furthermore, the image acquisition unit 201 performs an opening operation on the first target region I(z) within the slice including the patient coordinate z with a structural element having a size inversely proportional to the circularity defined by the following formula (13), thereby being able to identify and remove the locations protruding from the cylindrical shape within the spinal canal region information.

[0095]

Equation

[0096] By thus improving the accuracy of spinal canal segmentation, the image acquisition unit 201 can more appropriately perform the reading of spinal canal infiltration.

[0097] In improving the accuracy of spinal canal segmentation, the opening operation may be substituted with other spatial filtering processes. In this case, the size of the structural element generally corresponds to the size of the kernel of the spatial filter.

[0098] (Modification Example 2: Visualization of the Occupancy Rate of Extraskeletal Tumors) For the first embodiment, in the image acquisition unit 201, by setting the target region to the peripheral region of the vertebra (for example, a region within a predetermined distance R [mm] from the periphery of the vertebra) instead of the spinal canal, it is also possible to visualize the occupancy rate of the extraosseous tumor occurring around the vertebra. It is desirable to set the specific value of R so that it does not affect other organs such as the lungs around the vertebra. Specifically, about 5 mm is appropriate. Also, since the influence of beam hardening artifacts is small in extraosseous tumors, correction 1 may not be performed.

[0099] (Modification 3: Example other than the spinal canal) For the first embodiment, in the image acquisition unit 201, by setting the target region to the brain parenchyma instead of the spinal canal, it is also possible to visualize the lesions occurring in the brain parenchyma. In the brain parenchyma, the influence of beam hardening artifacts is more likely to occur closer to the skull. Therefore, it is desirable to calculate the reliability according to the distance from the skull or correct the reliability.

[0100] (Modification 4: Example other than comparison with the average value) For the first embodiment, in the first reliability estimation unit 202 and the second reliability estimation unit 203, in addition to the average pixel value of the spinal canal region, other statistics such as the median value or the mode value of the pixel values in the spinal canal region can also be used as the values to be compared with each pixel value.

[0101] (Modification 5: Example other than beam hardening artifacts) For the first embodiment, in the first reliability estimation unit 202 and the second reliability estimation unit 203, by adjusting the reference value μ, it is also possible to cope with the noise caused by metal artifacts. Specifically, since the artifacts caused by metal greatly reduce the pixel values around the metal compared to bone, the value of the reference value μ needs to be set smaller than in the case of bone.

[0102] <The Second Embodiment> As another embodiment according to the present application, a form in which the difference data is displayed corresponding to the reliability of the difference value can be considered.

[0103] FIG. 4 is a diagram showing an example of the configuration of the image processing apparatus 20 according to the second embodiment. The second embodiment is different from the first embodiment in that a differential data display function 11g is added. Specifically, the differential data display function 11g according to the second embodiment causes the display 14 to display the differential data calculated by the differential data calculation function 11f in correspondence with the reliability of the differential value calculated by the differential value reliability calculation function 11e. Here, the differential data display function 11g is an example of a differential data display unit. Hereinafter, the description will be centered around these.

[0104] FIG. 5 is a flowchart showing the processing steps executed by the image processing apparatus 20 according to the second embodiment. FIG. 6 is a data flow diagram showing all the steps of the image processing apparatus 20 according to the second embodiment.

[0105] The differential data display unit 206 takes as inputs the differential data output by the differential data calculation unit 205 and the reliability of the differential value output by the differential value reliability calculation unit 204, and displays the differential data on the display 14 in a display mode corresponding to the reliability of the differential value. In some cases, it may be paraphrased that the differential data display unit 206 takes as inputs the differential data output by the differential data calculation unit 205 and the reliability of the differential value output by the differential value reliability calculation unit 204, and displays the differential data on the display 14 in correspondence with the reliability of the differential value.

[0106] The processing from step S3010 to S3050 in FIG. 5 is the same as the processing from step S2010 to S2050 in FIG. 2 of the first embodiment, and the detailed description thereof will be omitted. Hereinafter, step S3060 will be described.

[0107] (Step S3060) In the differential data display unit 206, differential data such as a differential image, a fusion image, and occupancy rate data is displayed on the display 14 in a display mode in which the reliability of each differential value can be visually recognized by the user.

[0108] For example, when the differential data display unit 206 displays the fused image on the display 14, as shown in FIG. 15, it draws by changing the color of the infiltration site 122 or the infiltration site 123 in correspondence with the reliability of each differential value in the spinal canal 121. That is, for example, the differential data display unit 206 makes the infiltration site 123 with high reliability brighter than others, and draws the infiltration site 122 with low reliability affected by noise darker than others.

[0109] Also, for example, when the differential data display unit 206 displays the occupancy rate data on the display 14, the average reliability b of the differential values in the first target region I(z) within the slice including the patient coordinate z m is calculated by the following formula (14), and is displayed on the display 14 in association with the occupancy rate graph G idx as shown in the graph B of FIG. 16. idx

[0110]

Equation

[0111] Here, B in FIG. 16 idx represents a graph of the average reliability. On each axial plane of the target image, b m (z) representing the average reliability is a graph plotted along the Z axis of the image I sag . In FIG. 16, the meanings of the other symbols are the same as those of the symbols in FIG. 13.

[0112] As another display example, the differential data display unit 206 draws by changing the luminance value of the occupancy rate graph as shown in G'' of FIG. 16 in correspondence with the average reliability b m . That is, the differential data display unit 206 draws the occupancy rate graph at the location with low average reliability darker than other locations, such as P2. idx

[0113] As described above, by using the image processing apparatus 20 according to the second embodiment, the reliability of each differential value can be displayed so that the user can understand it, and the user can appropriately perform the reading using the differential value.

[0114] <Third Embodiment> As another embodiment according to the present application, a form using an inference device obtained by machine learning for estimating reliability can be considered.

[0115] FIG. 7 is a diagram showing an example of the configuration of the image processing apparatus 30 according to the third embodiment. The third embodiment is different from the first embodiment in that the first reliability estimation function 11c is replaced by the lesion degree estimation function 11h, and the second reliability estimation function 11d is replaced by the noise degree estimation function 11i. Here, the lesion degree estimation function 11h is an example of a lesion degree estimation unit, and the noise degree estimation function 11i is an example of a noise degree estimation unit. Hereinafter, the description will be centered on these.

[0116] FIG. 8 is a flowchart showing all the steps of the image processing apparatus 30 according to the third embodiment. Further, FIG. 9 is a data flow diagram showing all the steps of the image processing apparatus 30 according to the third embodiment.

[0117] Since the processing from step S4010 to step S4013 is the same as the processing from step S2010 to step S2013 in the first embodiment, the detailed description is omitted.

[0118] In FIG. 8, step S4020 and step S4030 are executed as sequential processing, but they may be executed as parallel processing.

[0119] (Step S4020) In the lesion degree estimation unit 302, an inference is performed on the first pixel value in the first region image using a pre-trained inference device, and the lesion degree b1 of the first region image is estimated. Here, the inference device can be constructed using a known machine learning method including deep learning and the like. In this case, it is assumed that cases with and without lesions in the spinal canal have been pre-trained using learning data. Further, as an input to the inference device, statistical values of the first target region in the first image, such as the average value, the mode value, the median value, etc. of the pixels in the spinal canal region, may be added. The lesion degree b1 calculated in this processing step calculates a value that is 1 when it is most likely to be a lesion and 0 when it is least likely to be a lesion, and calculates a value between 0 and 1 according to the degree.

[0120] (Step S4030) In the noise degree estimation unit 303, an inference is performed on the second pixel value in the second region image using a pre-trained inference device, and the noise degree b2 of the second region image is estimated. Here, the inference device can be constructed using a known machine learning method including deep learning and the like. In this case, it is assumed that cases with and without noise due to beam hardening artifacts, etc. in the spinal canal have been pre-trained using learning data. Further, as an input to the inference device, statistical values of the second target region in the second image may be added. The noise degree b2 calculated in this processing step calculates a value that is 1 when it is most likely to be noise and 0 when it is least likely to be noise, and calculates a value between 0 and 1 according to the degree.

[0121] (Step S4040) In the difference value reliability calculation unit 304, using the lesion degree b1 and the noise degree b2, for example, the reliability b of the difference value is calculated by the following formula (15). Note that as long as it is calculated based on the lesion degree b1 and the noise degree b2, the calculation formula of the difference value reliability is not limited to this. d is calculated.

[0122]

Equation

[0123] Here, since the higher the noise level b2 is, the lower the reliability of the difference value, the value subtracted from 1 is used in the above formula (15).

[0124] The difference value output from the image processing apparatus 30 is the result of multiplying the original difference value by the reliability bd of this difference value.

[0125] The difference value reliability calculation unit 304 may further correct the difference value reliability according to the intervertebral foramen level or the vertebral level, similarly to the first embodiment.

[0126] (Step S4050) The difference data calculation unit 305 calculates the pixel value of the difference image in the same manner as in the first embodiment. Further, the difference data calculation unit 305 may calculate a fusion image obtained by assigning a specified color to the pixel value of the difference image and superimposing it on the current image as difference data, similarly to the first embodiment. Further, the difference data calculation unit 305 may calculate the occupancy rate (occupancy rate data) g(z) within the slice including the patient coordinate z, similarly to the first embodiment.

[0127] The difference data calculation unit 305 outputs the difference image, the fusion image, and the occupancy rate data calculated as described above as difference data to an external PACS or the like.

[0128] As described above, according to the image processing apparatus 30 according to the third embodiment, it is possible to exclude the influence of noise and allow a user such as a doctor to perform reading using the difference without load. That is, by using the image processing apparatus 30 according to the third embodiment, it is possible to solve the problem that reading using the difference value cannot be appropriately performed due to the influence of noise.

[0129] <Other Embodiments> In the above-described embodiments, the parts to be identified are the spinal canal part and the part around the vertebra, but the parts to be identified are not limited to this.

[0130] Although the above embodiments have been described in detail, the present invention can be implemented, for example, as a system, apparatus, method, program, or recording medium (storage medium). Specifically, it may be applied to a system composed of a plurality of devices (for example, a host computer, an interface device, an imaging device, a web application, etc.), or it may also be applied to an apparatus composed of a single device.

[0131] Also, the term "processor" used in the description of the above embodiments means, for example, a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), or a circuit such as an application specific integrated circuit (ASIC), a programmable logic device (for example, a simple programmable logic device (SPLD), a complex programmable logic device (CPLD), and a field programmable gate array (FPGA)). Here, instead of storing the program in the storage circuit, the program may be directly incorporated into the circuit of the processor. In this case, the processor realizes its function by reading and executing the program incorporated in the circuit. Also, each processor of the present embodiment is not limited to being configured as a single circuit for each processor, and may be configured as a single processor by combining a plurality of independent circuits to realize its function.

[0132] Here, the image processing program executed by the processor is provided by being pre-installed in a ROM (Read Only Memory), a storage circuit, or the like. Note that this image processing program may be provided by being recorded on a non-transitory computer-readable storage medium such as a CD (Compact Disk)-ROM, an FD (Flexible Disk), a CD-R (Recordable), or a DVD (Digital Versatile Disk) in a file in a form installable or executable on these devices. Further, this image processing program may be stored on a computer connected to a network such as the Internet and provided or distributed by being downloaded via the network. For example, this image processing program is composed of modules including the above-described respective processing functions. As actual hardware, the CPU reads and executes a medical information processing program from a storage medium such as a ROM, whereby each module is loaded onto the main storage device and generated on the main storage device.

[0133] Also, in the above-described embodiments, each component of each illustrated device is conceptually functional and does not necessarily have to be physically configured as illustrated. That is, the specific form of the dispersion or integration of each device is not limited to that illustrated, and all or a part of it can be functionally or physically dispersed or integrated in any unit according to various loads, usage situations, and the like. Further, each processing function performed by each device can be realized in whole or in any part thereof by a CPU and a program analyzed and executed by the CPU, or can be realized as hardware by wired logic.

[0134] In addition, among the processes described in the above-described embodiments and modifications, all or part of the processes described as being automatically performed can be manually performed, or all or part of the processes described as being manually performed can be automatically performed by a known method. In addition, regarding the processing procedures, control procedures, specific names, and information including various data and parameters shown in the above documents and drawings, they can be arbitrarily changed unless otherwise specified.

[0135] Regarding the above embodiments, the following additional remarks are disclosed as one aspect and optional features of the invention.

[0136] (Supplementary Note 1) An image processing apparatus provided in one aspect of the present invention includes an acquisition unit that acquires a data sequence of reflected wave data collected in a plurality of frames in the time direction at one or a plurality of positions by ultrasonic transmission and reception, and from the data sequences of the plurality of frames, a Doppler processing unit that reduces clutter components derived from tissue and estimates first blood flow information, generates second blood flow information that is an image of scalar values of blood flow signals from the first blood flow information, detects a local maximum value of the second blood flow information, generates a scalar value integrated image by integrating the local maximum value or an image in which the local maximum value is emphasized, calculates an autocorrelation function in the frame direction of the first blood flow information at the position where the local maximum value is taken, generates an autocorrelation function integrated image by integrating the autocorrelation function, generates third blood flow information in which blood flow is color-coded from the scalar value integrated image and the autocorrelation function integrated image, and a super-resolution processing unit that causes the display unit to display the third blood flow information.

[0137] (Supplementary Note 2) The pixel value of the first pixel constituting the first region image is the first pixel value, the pixel value of the second pixel constituting the second region image is the second pixel value, and the first pixel and the second pixel may be pixels at corresponding positions.

[0138] (Supplementary Note 3) The first reliability estimation unit may estimate the reliability of the first pixel value based on the standard pixel value, the first pixel value, and the first reference value of the first target region in the first image including the first region image, or the second reliability estimation unit may estimate the reliability of the second pixel value based on the standard pixel value, the second pixel value, and the second reference value of the second target region in the second image including the second region image.

[0139] (Appendix 4) It may further include a difference data calculation unit that calculates difference data created from the difference value based on the first image, the second image, and the reliability of the difference value.

[0140] (Appendix 5) The difference value reliability calculation unit may correct the reliability of the difference value according to the vertebral level or intervertebral foramen level to which the first region image belongs.

[0141] (Appendix 6) It may further include a difference data display unit that causes the display unit to display the difference data corresponding to the reliability of the difference value.

[0142] (Appendix 7) The first image and the second image may be images captured by an X-ray CT (Computed Tomography) device.

[0143] (Appendix 8) The first reference value may be a value determined by the representative pixel value of the lesion that appears in the first target region, and the second reference value may be a value determined by the representative pixel value of the location where beam hardening artifacts occur.

[0144] (Appendix 9) When extracting the first target region from the first image including the first region image, the image acquisition unit may change the size of the kernel of the spatial filtering according to the distance between the spatial filtering target and the adjacent intersection hole.

[0145] (Appendix 10) When extracting the first target area from the first image including the first area image, the image acquisition unit may change the threshold value of the first pixel value regarded as the first target area according to the position in the body axis direction.

[0146] (Appendix 11) When extracting the first target area from the first image including the first area image, the image acquisition unit may change the size of the kernel of the spatial filtering according to the size of the neighborhood circularity of the spatial filtering target.

[0147] (Appendix 12) The second image may be an image captured at a time different from the first image.

[0148] (Appendix 13) An image processing method provided in one aspect of the present invention includes an image acquisition step of acquiring a first image including a first area image of a common subject captured, and a second image including a second area image corresponding to the first area image, a first reliability estimation step of estimating the reliability of the first pixel value based on the first pixel value and the first reference value constituting the first area image, a second reliability estimation step of estimating the reliability of the second pixel value based on the second pixel value and a second reference value different from the first reference value constituting the second area image, and a difference value reliability calculation step of obtaining the reliability of the difference value between the first pixel value and the second pixel value based on the reliability of the first pixel value and the reliability of the second pixel value.

[0149] (Appendix 14) A program provided in one aspect of the present invention is a program for causing a computer to execute the image processing method described in Appendix 13.

[0150] (Appendix 15) In another aspect of the present invention, an image processing apparatus provided includes an image acquisition unit that acquires a first image including a first region image of a common subject that has been imaged, and a second image including a second region image corresponding to the first region image; a lesion degree estimation unit that estimates the degree of lesion of the first region image using a first inference device based on first pixel values constituting the first region image; a noise degree estimation unit that estimates the degree of noise of the second region image using a second inference device different from the first inference device based on second pixel values constituting the second region image; and a difference value reliability calculation unit that calculates the reliability of a difference value between the first pixel values and the second pixel values based on the degree of lesion and the degree of noise.

[0151] (Appendix 16) In another aspect of the present invention, an image processing apparatus provided includes an image acquisition unit that acquires a first image including a first region image of a common subject that has been imaged and a second image different from the first image including a second region image corresponding to the first region image; and a difference value reliability calculation unit that acquires the reliability of a difference value between first pixel values constituting the first region image and second pixel values constituting the second region image according to the vertebral level or intervertebral foramen level to which the first region image belongs.

[0152] According to at least one of the embodiments described above, it is possible to allow a user such as a doctor to perform reading using a difference without load by excluding the influence of noise.

[0153] Although several embodiments have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the gist of the invention. These embodiments and their modifications are included in the scope and gist of the invention, and are also included in the invention described in the claims and its equivalent scope.

Description of Reference Numerals

[0154] 10, 20, 30 Image processing apparatus 11c First reliability estimation function 11d Second reliability estimation function 11e Difference value reliability calculation function

Claims

1. An image acquisition unit that acquires a first image including a first region image of a common subject that has been imaged, and a second image that is different from the first image and includes a second region image corresponding to the first region image; A first reliability estimation unit that estimates the reliability of the first pixel value based on the first pixel value and the first reference value that constitute the first region image; A second reliability estimation unit that estimates the reliability of the second pixel value based on the second pixel value that constitutes the second region image and a second reference value that is different from the first reference value; A difference value reliability calculation unit that acquires the reliability of the difference value between the first pixel value and the second pixel value based on the reliability of the first pixel value and the reliability of the second pixel value; An image processing apparatus comprising the above.

2. The image processing apparatus according to claim 1, wherein the pixel value of the first pixel that constitutes the first region image is the first pixel value, the pixel value of the second pixel that constitutes the second region image is the second pixel value, and the first pixel and the second pixel are pixels at corresponding positions.

3. The first reliability estimation unit estimates the reliability of the first pixel value based on the standard pixel value, the first pixel value, and the first reference value of a first target region in the first image including the first region image, or Or, The second reliability estimation unit estimates the reliability of the second pixel value based on the standard pixel value, the second pixel value, and the second reference value of a second target region in the second image including the second region image. The image processing apparatus according to claim 1.

4. The image processing apparatus according to claim 1, further comprising a difference data calculation unit that calculates difference data created from the difference value based on the first image, the second image, and the reliability of the difference value.

5. The image processing apparatus according to claim 1, wherein the difference value reliability calculation unit corrects the reliability of the difference value according to the vertebral level or the intervertebral foramen level to which the first region image belongs.

6. A difference data display unit that causes the display unit to display the difference data corresponding to the reliability of the difference value The image processing apparatus according to claim 4, further comprising the above.

7. The image processing apparatus according to claim 1, wherein the first image and the second image are images captured by an X-ray CT (Computed Tomography) apparatus.

8. The first reference value is a value determined by a representative pixel value of a lesion that appears in a first target region within the first image, and the second reference value is a value determined by a representative pixel value of a location where a beam hardening artifact occurs. The image processing apparatus according to claim 1.

9. When extracting a first target region from the first image including the first region image, the image acquisition unit changes the size of the kernel of the spatial filtering according to the distance between the spatial filtering target and the neighboring cross holes. The image processing apparatus according to claim 1.

10. When extracting a first target region from the first image including the first region image, the image acquisition unit changes the threshold value of the first pixel value regarded as the first target region according to the position in the body axis direction. The image processing apparatus according to claim 1.

11. When extracting a first target region from the first image including the first region image, the image acquisition unit changes the size of the kernel of the spatial filtering according to the size of the circularity of the vicinity of the spatial filtering target. The image processing apparatus according to claim 1.

12. The second image is an image captured at a time different from that of the first image. The image processing apparatus according to claim 1.

13. An image acquisition step of acquiring a first image including a first region image of a common subject captured, and a second image including a second region image corresponding to the first region image; A first reliability estimation step of estimating the reliability of the first pixel value based on the first pixel value constituting the first region image and the first reference value; A second reliability estimation step of estimating the reliability of the second pixel value based on the second pixel value constituting the second region image and a second reference value different from the first reference value; A differential value reliability calculation step of obtaining the reliability of the differential value between the first pixel value and the second pixel value based on the reliability of the first pixel value and the reliability of the second pixel value; An image processing method comprising:

14. A program for causing a computer to execute the image processing method according to claim 13.

15. An image acquisition unit that acquires a first image including a first region image of a common subject captured, and a second image including a second region image corresponding to the first region image; A lesion degree estimation unit that estimates the degree of lesion of the first region image using a first inference device based on the first pixel value constituting the first region image; A noise level estimation unit that estimates the noise level of the second region image using a second inference device different from the first inference device based on the second pixel values that make up the second region image; A difference value reliability calculation unit that calculates the reliability of the difference value between the first pixel value and the second pixel value based on the degree of lesion and the noise level; An image processing apparatus comprising:

16. An image acquisition unit that acquires a first image including a first region image of a common subject that has been imaged, and a second image different from the first image including a second region image corresponding to the first region image; A difference value reliability calculation unit that acquires the reliability of the difference value between the first pixel value that makes up the first region image and the second pixel value that makes up the second region image according to the vertebral level or intervertebral foramen level to which the first region image belongs; An image processing apparatus comprising:

Images