JP2017131454A - Medical image display system, medical image display program and medical image display method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accelerate positioning on a medical image.SOLUTION: A medical image display system comprises: a feature point extraction part for extracting a feature point used for positioning with other medical image, from a medical image; a density threshold processing part for extracting a grayscale pixel area of a pixel having a density value in a prescribed density range from the medical image; an edge detection part for detecting an edge pixel area from the medical image; a determination part for specifying a position of the edge pixel area included in the inside of the grayscale pixel area extracted from the medical image, in the grayscale pixel area, and for determining whether or not the grayscale pixel area is a disease area; and a positioning processing part performing positioning of the medical image and other medical image, using the feature point included in the grayscale pixel area which is determined that it is not the disease area by the determination part, out of the feature points extracted from the medical image.SELECTED DRAWING: Figure 12

Description

  The present invention relates to a medical image display system, a medical image display program, and a medical image display method.

  In the medical field, in order to judge the progress of a patient's disease, an interpreting doctor may comparatively interpret medical images such as CT (Computed Tomography) images taken at different times using a medical image display system. . At this time, when the part of the patient to be comparatively interpreted is the lung or the like, each region of the CT image includes position fluctuations accompanying the patient's heartbeat and respiration. For this reason, in a medical image display system, when displaying a predetermined region in a CT image, feature points (for example, blood vessels, bronchial bifurcations) are extracted, and alignment is performed so that the positions of the feature points match. As a result, the variation of the position is corrected for the region to be compared. Thereby, the interpretation doctor can easily perform comparative interpretation for each region even in the case of a CT image of a region such as the lung.

JP 2013-141603 A

  Here, when performing matching as described above, there is a problem that it takes time to perform alignment if the number of extracted feature points is large. In particular, in the case of a system that performs registration using all the extracted feature points, the extracted feature points include feature points that are likely to fail in matching, and the alignment takes time. It is a factor.

  An object of one aspect is to speed up alignment in medical images.

According to one aspect, a medical image display system includes:
A feature point extraction unit for extracting a feature point used for alignment with another medical image from the medical image;
From the medical image, a density threshold processing unit that extracts a gray pixel region including pixels having density values within a predetermined density range;
From the medical image, an edge detection unit that detects an edge pixel region;
Determination that determines whether or not the gray pixel area is a disease area by specifying the position of the edge pixel area included in the gray pixel area extracted from the medical image in the gray pixel area And
Among the feature points extracted from the medical image, the medical image and the other medical image are used by using the feature point included in the gray pixel region determined not to be a disease region by the determination unit. And an alignment processing unit for performing the alignment.

  Positioning in a medical image can be speeded up.

It is a figure which shows an example of CT image imaging system. It is a figure which shows the hardware constitutions of a medical image display system. It is a 1st figure which shows the relationship between the processing content of the diagnosis assistance part in a medical image display system, the operation content of a radiogram interpretation doctor, and the display content of a parallel display screen. It is a 2nd figure which shows the relationship between the processing content of the diagnostic assistance part in a medical image display system, the operation content of an image interpretation doctor, and the display content of a parallel display screen. It is a figure which shows an example of the information stored in image DB. It is a figure which shows the detail of a function structure of a 2nd registration part. It is a figure for demonstrating the detail of the processing content of a feature point extraction part. It is the figure which showed the relationship between CT value, the density value in a CT image, and a patient's tissue. It is a 1st figure for demonstrating the process which determines whether the grayscale pixel area extracted in the feature point filtering part is a disease area | region. It is a 2nd figure for demonstrating the process which determines whether the grayscale pixel area extracted in the feature point filtering part is a disease area | region. It is a 3rd figure for demonstrating the process which determines whether the grayscale pixel area extracted in the feature point filtering part is a disease area | region. It is a 1st figure which shows the function structure of a feature point filtering part. It is a 1st flowchart of a feature point filtering process. It is a figure which shows the image before and behind a feature point filtering process. It is a 2nd figure which shows the detail of a function structure of a feature point filtering part. It is a figure which shows the reliability calculated about each feature point. It is a 2nd flowchart of a feature point filtering process.

  Each embodiment will be described below with reference to the accompanying drawings. In addition, in this specification and drawing, about the component which has the substantially same function structure, the duplicate description is abbreviate | omitted by attaching | subjecting the same code | symbol.

  In the following embodiments, “feature points effective for CT image alignment” are included, for example, in a region (normal region) that does not change with time or hardly changes with time between CT images. Refers to a feature point. This is because the feature points included in the normal region are highly likely to be successfully matched. Furthermore, in the following embodiments, “feature points that are not effective for CT image alignment” refer to feature points that are included in a disease region that is an initial state of cancer, such as a ground glass-like shadow, for example. . This is because the disease region has a high probability of changing with time as the disease progresses, and the feature points included in the disease region are likely to fail in matching.

[First Embodiment]
First, a CT (Computed Tomography) image capturing system including the medical image display system according to the first embodiment will be described. FIG. 1 is a diagram illustrating an example of a CT image imaging system.

  The CT image capturing system 100 includes a CT apparatus 110, a medical image display system 120, and an image database (hereinafter, the database is abbreviated as DB) 130. The CT apparatus 110 and the medical image display system 120 are connected via a wiring 111, and various types of data are transmitted and received between the both apparatuses. In addition, the medical image display system 120 and the image DB 130 are connected via a wiring 121, and various types of data are transmitted and received between both devices.

  The CT apparatus 110 scans the inside of a patient using radiation or the like, and generates a CT image which is a slice image of the patient by processing using a computer (hereinafter, such processing is referred to as “CT image”). Called "shoot"). The CT apparatus 110 transmits the captured CT image to the medical image display system 120.

  The medical image display system 120 stores CT images taken by the CT apparatus 110 in the connected image DB 130. A diagnostic support program is installed in the medical image display system 120, and the medical image display system 120 functions as the diagnostic support unit 140 when the diagnostic support program is executed by a computer.

  The image DB 130 receives CT images taken by the CT apparatus 110 via the medical image display system 120 and stores them separately for each of a plurality of CT images (imaging groups) taken at the same time.

  The diagnosis support unit 140 is used when an interpreting doctor interprets a CT image stored in the image DB 130. The diagnosis support unit 140 displays, for example, CT images taken at different times in parallel so that the interpretation doctor can perform comparative interpretation. In the following, one of the CT images displayed in parallel (for example, a CT image taken before the lapse of a predetermined period) is referred to as a “comparison source CT image”, and the other (for example, a CT image taken after the lapse of a predetermined period). Image) is referred to as a “comparison target CT image”.

  The diagnosis support unit 140 enlarges and displays an image of a predetermined region (ROI: Region of interest) including the position designated by the interpretation doctor in the comparison source CT image on the enlarged display screen. Further, the diagnosis support unit 140 extracts an image of a corresponding area corresponding to a predetermined area including the designated position from the comparison target CT image, and displays the enlarged image on the enlarged display screen.

  The diagnosis support unit 140 includes a first registration unit 141, a second registration unit 142, and a display control unit 143 in order to execute these processes.

  The first registration unit 141 is realized, for example, by executing a first registration program by a computer. When the first registration unit 141 displays CT images taken at different times in parallel, the first registration unit 141 corrects misalignment between the CT images by affine transformation, thereby performing global alignment between the CT images. I do.

  For example, the second registration unit 142 is realized by a computer executing a second registration program that is an example of a medical image display program. The second registration unit 142 extracts an image of a predetermined region including the position designated by the interpretation doctor, performs local alignment by performing a conversion process on the comparison target CT image, and performs comparison with the comparison target CT image. Extract the corresponding region image. Although various processes are included in the conversion process, in the first embodiment, the conversion process refers to a parallel movement, and in the following, the correspondence extracted from the comparison target CT image by performing the conversion process. The image of the region is referred to as “image with local alignment”.

  The display control unit 143 is realized, for example, by executing a display program by a computer. The display control unit 143 controls to display the comparison source CT image selected by the interpretation doctor, and controls to enlarge and display an image of a predetermined area including the position designated by the interpretation doctor. In addition, the display control unit 143 performs control so that the image extracted by the second registration unit 142 and subjected to local alignment is enlarged and displayed on the enlarged display screen.

  Next, the hardware configuration of the medical image display system 120 will be described. FIG. 2 is a diagram illustrating a hardware configuration of the medical image display system. As shown in FIG. 2, the medical image display system 120 includes a CPU (Central Processing Unit) 201, a ROM (Read Only Memory) 202, and a RAM (Random Access Memory) 203. The medical image display system 120 includes an auxiliary storage device 204, a connection device 205, a display device 206, an operation device 207, and a drive device 208. Each unit of the medical image display system 120 is connected to each other via a bus 209.

  The CPU 201 is a computer that executes various programs (for example, a first registration program, a second registration program, a display program, etc.) stored in the auxiliary storage device 204.

  The ROM 202 is a nonvolatile memory. The ROM 202 functions as a main storage unit that stores various programs, data, and the like necessary for the CPU 201 to execute various programs stored in the auxiliary storage device 204. Specifically, the ROM 202 stores a boot program such as a basic input / output system (BIOS) or an extensible firmware interface (EFI).

  The RAM 203 is a volatile memory, and includes a DRAM (Dynamic Random Access Memory), an SRAM (Static Random Access Memory), and the like. The RAM 203 is a main storage unit that provides a work area that is expanded when various programs stored in the auxiliary storage device 204 are executed by the CPU 201.

  The auxiliary storage device 204 is a computer-readable storage device that records various programs installed in the medical image display system 120 and data generated by executing the various programs.

  The connection device 205 is connected to the CT device 110 and the image DB 130, and transmits and receives various data between the CT device 110 and the image DB 130. The display device 206 displays the CT image stored in the image DB 130 on the parallel display screen under the control of the display control unit 143. The operation device 207 accepts various operations performed on the medical image display system 120 by the image interpretation doctor.

  The drive device 208 is a device for setting the recording medium 210. The recording medium 210 here includes a medium for recording information optically, electrically or magnetically, such as a CD-ROM, a flexible disk, a magneto-optical disk or the like. The recording medium 210 also includes a semiconductor memory that electrically records information, such as a ROM and a flash memory.

  The various programs stored in the auxiliary storage device 204 are installed by, for example, setting the distributed recording medium 210 in the drive device 208 and reading the various programs recorded in the recording medium 210 by the drive device 208. Is done. Alternatively, various programs stored in the auxiliary storage device 204 are installed by being downloaded from the network via the connection device 205.

  Next, the relationship between the processing content of the medical image display system 120 functioning as the diagnosis support unit 140, the operation content of the interpretation doctor at that time, and the parallel display screen displayed on the display device 206 of the medical image display system 120 will be described. To do.

  3 and 4 are first and second diagrams showing the relationship among the processing contents of the diagnosis support unit, the operation contents of the interpretation doctor, and the display contents of the parallel display screen in the medical image display system.

  By the medical image display system 120 functioning as the diagnosis support unit 140, processing by the display control unit 143 is started, and the display device 206 displays parallel images for displaying CT images taken at different times in parallel. A screen 300 is displayed (see FIG. 3). The parallel display screen 300 is provided with a function for an interpreting doctor to select an imaging group of a predetermined part (here, lungs) imaged at a predetermined time for a predetermined patient as a comparison source CT image group. Yes.

  When the comparison source CT image group is selected by the interpretation doctor, the display control unit 143 reads out the selected comparison source CT image group from the image DB 130. Further, when a predetermined comparison source CT image (here, file name = “ImageA015”) is designated by the interpretation doctor from the selected comparison source CT image group, the display control unit 143 displays the designated comparison source CT image. Control is performed so that the CT image is displayed on the parallel display screen 300.

  The parallel display screen 300 further has a function for the image interpretation doctor to select, as a comparison target CT image group, an imaging group of the same part of the same patient captured at different times to be compared with the comparison source CT image. Is provided. Specifically, a function is provided for selecting a comparison target CT image group by inputting a patient ID, an imaging date and time, an imaging region (in this case, a lung), and the like.

  When the patient name, imaging date / time, imaging region, and the like are input by the interpretation doctor, the display control unit 143 reads the imaging group specified by the input information from the image DB 130 as a comparison target CT image group. Further, when a predetermined comparison target CT image (here, file name = “ImageB018”) is designated by the interpretation doctor from the read comparison target CT image group, the display control unit 143 displays the designated comparison target CT image. Are controlled to be displayed on the parallel display screen 300.

  At this time, in the diagnosis support unit 140, the first registration unit 141 starts processing, and global alignment is performed by correcting each read CT image using affine transformation such as rotation and parallel movement. (See FIG. 4). The first registration unit 141 performs global positioning on the entire CT image, so that the global positional deviation between the comparison source CT image and the comparison destination CT image is eliminated.

  When the global alignment is completed, the image interpretation doctor can designate the position of the tumor portion F in the comparison source CT image displayed on the parallel display screen 300. When the interpreting doctor designates the position of the tumor part F on the parallel display screen 300, the display control unit 143 displays an image of a predetermined region (ROI: Region of interest) 401 including the designated position of the tumor part F as a comparison source. Control is performed so that the enlarged display is performed on the enlarged display screen on the CT image.

  When the image of the predetermined area 401 is enlarged and displayed, the second registration unit 142 performs local alignment by performing a conversion process on the comparison target CT image. Thereby, the second registration unit 142 extracts an image of the corresponding region including the position of the tumor portion F ′ corresponding to the tumor portion F (an image on which local alignment has been performed). In addition, the second registration unit 142 notifies the display control unit 143 of the image of the corresponding area.

  The display control unit 143 controls to enlarge and display the image of the corresponding region 402 notified from the second registration unit 142 on the enlarged display screen on the comparison target CT image. Thereby, on the parallel display screen 300, an image of the corresponding region 402 including the position of the tumor portion F ′ corresponding to the tumor portion F is displayed as an image on which the local alignment has been performed.

  As described above, according to the medical image display system 120, when the interpretation doctor designates the position of the tumor portion F in the comparison source CT image, the image of the predetermined region 401 can be enlarged and displayed. Further, the image of the corresponding region 402 can be automatically extracted from the comparison target CT image and can be enlarged and displayed on the enlarged display screen. As a result, the interpreting physician can easily perform comparative interpretation of the corresponding positions between the CT images included in the imaging groups taken at different times, and easily diagnose how the tumor has changed. Can do.

  Next, the image DB 130 will be described. FIG. 5 is a diagram illustrating an example of information stored in the image DB. As shown in FIG. 5, the information stored in the image DB 130 is classified and managed for each patient, and FIG. 5 shows an example of information about a patient with patient ID = “xxx”.

  As shown in FIG. 5, the information items include “imaging date / time”, “imaging region”, “series name”, and “imaging group”. Information on the date and time when the CT image was captured is stored in the “imaging date and time”. Information on a specific part to be imaged is stored in “imaging part”. The “series name” stores a series name for identifying a series composed of a plurality of CT images obtained by imaging. The “imaging group” stores file names of a plurality of CT images obtained by imaging.

  In the example of FIG. 5, the series name including the CT images of ImageA001 to ImageA030 obtained by performing imaging on the imaging region = “lung” at imaging date = “H26.2.5” = “series A”. It shows that the series is stored in the image DB 130. Further, the series of the series name = “series B” including the CT images of ImageB001 to ImageB030 obtained by performing imaging on the imaging region = “lung” at imaging date = “H26.8.3” is the image DB 130. Is stored.

  Note that the dotted line in FIG. 5 indicates that the CT image of “Image A015” is designated by the interpretation doctor as the comparison source CT image. Further, the CT image of “ImageB018” is designated by the interpretation doctor as a comparison destination CT image.

  Next, functions of each unit of the diagnosis support unit 140 will be described. In the following, description of the functions of the first registration unit 141 and the display control unit 143 is omitted, and the function of the second registration unit 142 is mainly described.

  As described above, when the global alignment is completed, the overall position variation between the comparison source CT image and the comparison destination CT image is corrected, while the local position caused by the patient's heartbeat and respiration is corrected. Variations in general position remain. For this reason, when the image of the corresponding region 402 corresponding to the predetermined region 401 including the position of the tumor portion F designated by the image interpretation doctor is enlarged and displayed, the second registration unit 142 corrects the local position variation. The movement vector for calculating the position is calculated and local alignment is performed. The movement vector is a vector indicating the direction and size of the position change (that is, local position change) of the image in the corresponding region of the comparison target CT image with respect to the image of the predetermined region 401 of the comparison source CT image. . Based on the calculated movement vector, the second registration unit 142 performs a conversion process for translating the corresponding region of the comparison target CT image, thereby extracting an image on which local alignment has been performed. Thereby, the second registration unit 142 can extract the image of the corresponding region 402 from the comparison target CT image.

  The functional configuration of the second registration unit 142 that performs such processing will be described in detail with reference to FIG. FIG. 6 is a diagram illustrating details of a functional configuration of the second registration unit.

  As shown in FIG. 6, the second registration unit 142 includes a CT image input unit 601, a region extraction unit 602, a feature point extraction unit 603, a feature point filtering unit 604, a feature point matching unit 605, a totaling unit 606, a transformation. Part 607 and output part 608. Note that the feature point matching unit 605, the totaling unit 606, and the conversion unit 607 are collectively referred to as a local alignment processing unit 610.

  The CT image input unit 601 inputs the comparison source CT image and the comparison destination CT image that have been globally aligned by the first registration unit 141.

  The area extraction unit 602 extracts an image of a predetermined area 401 including the position of the tumor portion F designated by the image interpretation doctor from the input comparison source CT image. Further, the region extraction unit 602 extracts an image of a region corresponding to the predetermined region 401 from the input comparison target CT image. The image extracted by the region extraction unit 602 is input to the feature point extraction unit 603 and the conversion unit 607.

  The feature point extraction unit 603 extracts feature points for each image extracted by the region extraction unit 602. A feature point is, for example, a pixel (or pixel group) at a position corresponding to a blood vessel or bronchus branch, an end of a blood vessel or bronchus, a boundary surface of an organ, or the like in the image, and is extracted by the region extraction unit 602. It is used when performing local registration between images. Details of the processing content of the feature point extraction unit 603 will be described later.

  The feature point filtering unit 604 classifies the feature points extracted by the feature point extraction unit 603 into feature points that are effective for local alignment and feature points that are not effective for local alignment. In addition, the feature point filtering unit 604 outputs, to the feature point matching unit 605, feature points that are effective for local alignment among the classified feature points, and feature points that are not effective for local alignment. Is excluded.

  In this way, the feature point filtering unit 604 excludes feature points that are not effective for local alignment, so that the feature point matching unit 605 performs matching between the extracted feature points. The number of can be reduced. As a result, matching can be speeded up (that is, positioning can be speeded up).

  Further, the feature point filtering unit 604 outputs feature points effective for local alignment, so that the feature point matching unit 605 performs matching using the feature points effective for local alignment. It can be carried out. As a result, the alignment accuracy in the local alignment processing unit 610 can be improved. Details of the processing content of the feature point filtering unit 604 will be described later.

  The feature point matching unit 605 performs matching using the feature points output from the feature point filtering unit 604, recognizes the correspondence between the feature points, and calculates a movement vector between the corresponding feature points.

  The totaling unit 606 totalizes the movement vectors between corresponding feature points calculated by the feature point matching unit 605. Thereby, the totaling unit 606 calculates one movement vector used for conversion processing for extracting an image on which local alignment has been performed from the comparison target CT image.

  The conversion unit 607 extracts an image that has been subjected to local alignment from the comparison target CT image by performing conversion processing using one movement vector calculated by the totaling unit 606.

  The output unit 608 outputs the image of the predetermined region 401 and the image subjected to local alignment (image of the corresponding region 402) to the display control unit 143.

  Next, processing contents of the feature point extraction unit 603 and the feature point filtering unit 604 among the units included in the second registration unit 142 will be described in more detail.

  First, the processing content of the feature point extraction unit 603 will be described in detail with reference to FIG. FIG. 7 is a diagram for explaining details of processing contents of the feature point extraction unit.

  The feature point extraction unit 603 extracts feature points in the image using an extraction method called FAST (Feature from Accelerated Segment Test). FIG. 7A is a diagram for explaining a feature point extraction method by FAST.

  As shown in FIG. 7A, the feature point extraction unit 603 determines whether each pixel in the image of the region extracted by the region extraction unit 602 corresponds to the feature point or not. Sixteen pixels (pixels 701 to 716) on the surrounding circumference are acquired.

  In addition, the feature point extraction unit 603 determines that n or more pixels in the 16 acquired pixels (pixels 701 to 716) are continuously brighter than the target pixel P by the density threshold t or the density. When the pixel is darker than the threshold t, it is determined that the target pixel P corresponds to the feature point. The density threshold t is a predetermined value.

  The feature point extraction unit 603 extracts feature points by performing the same processing for all the pixels in the image of the region extracted by the region extraction unit 602.

  FIG. 7B shows an image of the predetermined area 401 extracted by the area extraction unit 602 (before feature point extraction). In FIG. 7B, black lines indicate blood vessels and bronchi. 7C, the feature point extraction unit 603 extracts feature points from the image of the predetermined region 401 extracted by the region extraction unit 602, and places a cross mark at the position of the extracted feature points. It shows a state.

  As shown in FIG. 7C, the feature point extraction unit 603 can extract a branch portion of a blood vessel or a bronchus (black line in the drawing), a blood vessel, an end portion of the bronchus, or the like as a feature point.

  Next, the processing content of the feature point filtering unit 604 will be described in detail with reference to FIGS. The feature point filtering unit 604 classifies the feature points extracted by the feature point extraction unit 603 into feature points that are effective for local alignment and feature points that are not effective. It is determined whether it is included in the disease area.

  For this reason, the feature point filtering unit 604 first performs density threshold processing on the image of the region extracted by the region extracting unit 602 to extract a gray pixel region including a disease region. Then, the feature point filtering unit 604 determines whether or not the extracted gray pixel region is a disease region (whether it is a disease region or a normal region).

  FIG. 8 is a diagram showing the relationship between CT values, density values in CT images, and patient tissues. The density value of each pixel in the CT image is obtained by converting the CT value detected when the CT apparatus 110 images a patient into, for example, 4096 gradations. The CT value detected by the CT apparatus 110 represents the degree of X-ray absorption of the substance, and is defined so that the CT value of water is 0 [HU] and the CT value of air is −1000 [HU]. Yes.

  In the example of FIG. 8, the density value when CT value = −1000 [HU] is 0, and the density value when CT value = 2000 [HU] is 4096. In this case, for example, 437 to 983 is in the range of −280 [HU] to −680 [HU] as the range of the concentration value of the disease region including the initial state of cancer such as ground glass-like shadow. Hereinafter, a disease region including an initial state of cancer such as a ground glass-like shadow is referred to as a “GGO (Ground Glass Opacity) region”.

  As described above, the feature point filtering unit 604 can extract the light and dark pixel region including the GGO region by performing the density threshold processing for extracting the pixel having the density value within the predetermined density range.

  On the other hand, as shown in FIG. 8, the density value of the GGO region is substantially equal to the density value of a thin blood vessel, bronchus, etc., and when performing the concentration threshold processing for extracting the GGO region, the thin blood vessel and bronchus are also extracted. Will be. Note that thin blood vessels, bronchi, and the like are extracted both inside the GGO region and inside normal lung regions other than the GGO region.

  Therefore, the feature point filtering unit 604 uses the positional relationship between the gray pixel area and the thin blood vessels, bronchi, and the like in determining whether the gray pixel area extracted by performing the density threshold process is a GGO area. . Specifically, the feature point filtering unit 604 performs density threshold processing by specifying the position of the contour of the gray pixel area and the position of a thin blood vessel or bronchus included in the gray pixel area. It is determined whether or not the extracted gray pixel area is a GGO area.

  Therefore, the feature point filtering unit 604 separately performs edge detection processing for detecting thin blood vessels, bronchi, and the like on the image of the region extracted by the region extracting unit 602. Thereby, the feature point filtering unit 604 can extract thin blood vessels, bronchi, and the like as edge pixel regions. In many cases, the GGO region gradually changes from the density value (low density value) of the surrounding normal tissue to the density value (high density value) of the abnormal tissue, and the outline is unclear. Therefore, no edge is detected in the outline of the GGO region.

  9 to 11 are first to third diagrams for describing processing for determining whether or not the grayscale pixel region extracted by the feature point filtering unit is a GGO region.

  FIG. 9A shows an image of a predetermined area 401 among the area images extracted by the area extraction unit 602. FIG. 9B shows a binary image obtained by performing edge detection processing on the image of the predetermined area 401 and extracting the edge pixel area. As shown in FIG. 9B, by performing edge detection processing on the image of the predetermined region 401, a pixel group of blood vessels and edge portions of bronchi 901 to 906 is extracted as an edge pixel region.

  FIG. 9C shows a density threshold value process (predetermined range (−280 to −680)) for the image of the predetermined area 401 among the image of the area extracted by the area extracting unit 602. 2 shows a binary image obtained by performing pixel extraction processing and extracting light and dark pixel regions. As shown in FIG. 9C, by performing density threshold processing on the image of the predetermined area 401, blood vessels, bronchi 901 to 904, and a GGO area 910 in the normal area are extracted as light and dark pixel areas.

  Further, FIG. 9D shows a state in which FIG. 9B and FIG. 9C are superimposed (for convenience of explanation, the edge pixel region is white in FIG. 9D). Shown). As shown in FIG. 9D, the edge pixel area in the gray pixel area corresponding to the normal area and the edge pixel area in the gray pixel area corresponding to the GGO area are positioned in the contour of the gray pixel area. And the position of the edge pixel region are greatly different.

  Specifically, the relationship between the position of the contour of the gray pixel area and the position of the edge pixel area included in the gray pixel area is summarized as shown in FIG.

  As shown in FIG. 10, in the case of a blood vessel or bronchus 901 in a normal region, the distance from the contour of the light and dark pixel region (black) to the edge pixel region (white) is short. On the other hand, in the case of the blood vessels and bronchi 905 and 906 in the GGO region 910, the distance from the contour of the light and dark pixel region (black) to the edge pixel region (white) is long.

  As described above, in determining whether the grayscale pixel area extracted by performing the density threshold processing is a GGO area, the distance from the position of the outline of the grayscale pixel area to the position of the edge pixel area is used. It is effective.

  FIG. 11 shows a calculation method for calculating the distance from the position of the contour of the gray pixel area to the position of the edge pixel area. In calculating the distance from the position of the contour of the gray pixel area to the position of the edge pixel area, the feature point filtering unit 604 performs initialization processing, distance conversion processing by raster scanning, and distance conversion processing by reverse raster scanning. Execute.

  Specifically, the feature point filtering unit 604 converts the image subjected to the density threshold processing into a first distance conversion image as initialization processing. Here, the pixel values of the pixel (i, j) in the image subjected to the density threshold processing and the first distance conversion image are set to p (i, j) and d1 (i, j), respectively. When the pixel value p (i, j) is 1, for each pixel (i, j), the feature point filtering unit 604 sets the pixel value d1 (i, j) to an arbitrary maximum value (however, the maximum value is Greater than 1). The feature point filtering unit 604 sets 0 to the pixel value d1 (i, j) when the pixel value p (i, j) is 0.

  Subsequently, the feature point filtering unit 604 executes distance conversion processing by raster scanning. The feature point filtering unit 604 performs raster scanning from the upper left end to the lower right end using the mask 1101 shown in FIG. 11A with respect to the first distance conversion image generated as described above. The calculation of Expression (1) is performed. Thereby, the feature point filtering unit 604 converts the first distance-converted image into the second distance-converted image. Here, the pixel value of the pixel (i, j) in the second distance conversion image is assumed to be d2 (i, j).

Further, the feature point filtering unit 604 performs reverse raster using the mask 1102 illustrated in FIG. 11A on the pixel value d2 (i, j) of the pixel (i, j) in the second distance conversion image. A distance conversion process is performed by scanning. The feature point filtering unit 604 performs the following equation (2) while performing reverse raster scanning from the lower right end to the upper left end of the second distance conversion image generated as described above. Thereby, the feature point filtering unit 604 converts the second distance conversion image into the third distance conversion image. Here, the pixel value of the pixel (i, j) in the third distance conversion image is d (i, j).

The feature point filtering unit 604 executes these processes to calculate the distance d from the contour of the light and dark pixel area to each pixel included in the light and dark pixel area.

  FIG. 11B shows a third distance conversion image for the grayscale pixel region corresponding to the GGO region 910. In the example of FIG. 11B, the distance from the contour of the gray pixel area for each pixel included in the gray pixel area corresponding to the GGO area 910 is expressed by a density value (the higher the density value, the distance is Long and low density values indicate a short distance).

  FIG. 11C shows the distance from the contour of the light and dark pixel region of the pixels 1111 and 1112 extracted as the edge pixel region among the pixels included in the light and dark pixel region corresponding to the GGO region 910. is there.

As shown in FIG. 11C, the feature point filtering unit 604 includes a distance d _1111a from the position of the upper left corner when viewed from the pixel 1111 to the pixel 1111 and the lower right corner of the gray pixel area. The distance d _1111b from the position to the pixel 1111 is calculated. Note that the feature point filtering unit 604 obtains the shorter one of the two calculated distances (d _1111a ) as the distance from the contour of the grayscale pixel region to the pixel 1111.

The feature point filtering unit 604 also includes a distance d _1112a from the position of the upper left corner of the contour of the gray pixel region corresponding to the GGO region 910 to the pixel 1111 when viewed from the pixel 1112, and the position of the contour of the lower right corner. To a distance d _1112b from the pixel 1112 is calculated. Note that the feature point filtering unit 604 _acquires the shorter distance (d _1112a ) of the two calculated distances as the distance from the contour of the gray pixel region to the pixel 1112.

  In the example of FIG. 11, a case has been shown in which the distance from the contour of the light and dark pixel region is acquired for two pixels among the pixels extracted as the edge pixel region. However, the feature point filtering unit 604 obtains the distance from the contour of the light and dark pixel region for all the pixels forming the edge pixel region included in the light and dark pixel region.

  Further, in the example of FIG. 11, the case where the distance is acquired for the gray pixel area corresponding to the GGO area 910 has been described. However, the feature point filtering unit 604 performs the gray pixel area corresponding to the blood vessels and bronchi 901 to 904. Performs the same processing.

  Note that the feature point filtering unit 604 generates a histogram based on the distances acquired for all the pixels forming the edge pixel region included in the gray pixel region, and calculates the distance of the mode value in the gray pixel region. Extract. Then, the feature point filtering unit 604 calculates the extracted mode value as a representative value of the distance from the contour of the gray pixel area of the edge pixel area included in the gray pixel area.

  Alternatively, the feature point filtering unit 604 calculates an average value of the distances acquired for all the pixels forming the edge pixel region included in the gray pixel region. Then, the feature point filtering unit 604 calculates the calculated average value as a representative value of the distance from the contour of the gray pixel area of the edge pixel area included in the gray pixel area.

  Alternatively, the feature point filtering unit 604 calculates the minimum value or the maximum value among the distances acquired for all the pixels forming the edge pixel region included in the gray pixel region. Then, the feature point filtering unit 604 calculates the calculated maximum value or minimum value as a representative value of the distance from the contour of the gray pixel area of the edge pixel area included in the gray pixel area.

  Next, a functional configuration of the feature point filtering unit 604 that realizes the processing contents will be described. FIG. 12 is a first diagram illustrating a functional configuration of the feature point filtering unit. As shown in FIG. 12, the feature point filtering unit 604 includes an edge detection unit 1201, a density threshold processing unit 1202, a labeling unit 1203, a minute point removal unit 1204, a distance calculation unit 1205, a determination unit 1206, and a feature point output unit. 1207.

  The edge detection unit 1201 performs an edge detection process for detecting pixels having a density gradient within a predetermined range on the image of the region extracted by the region extraction unit 602, and outputs the edge pixel region to the distance calculation unit 1205 ( (See FIG. 9B). The edge detection unit 1201 performs an edge detection process using an existing method such as Canny Edge Detector, for example.

  The density threshold processing unit 1202 extracts a pixel having a density value of density threshold processing (for example, within a predetermined density range (−280 to −680)) from the image of the region extracted by the region extraction unit 602. )I do. Further, the density threshold processing unit 1202 outputs the density pixels extracted by performing the density threshold processing to the labeling unit 1203 (see FIG. 9C).

  The labeling unit 1203 collects the output gray pixels from the density threshold processing unit 1202 for each region to form a gray pixel region, and performs labeling on each gray pixel region.

  The minute point removal unit 1204 removes the grayscale pixel area having a predetermined size or less from the grayscale pixel areas labeled by the labeling unit 1203, and calculates the distance of the grayscale pixel area other than the removed grayscale pixel area. To the unit 1205.

  The distance calculation unit 1205 acquires an edge pixel region from the edge detection unit 1201, and acquires a labeled grayscale pixel region from the labeling unit 1203. Further, the distance calculation unit 1205 superimposes the edge pixel area and the gray pixel area (see FIG. 9D). Further, the distance calculation unit 1205 calculates a representative value of the distance from the contour of the light and dark pixel area of the edge pixel area included in the light and dark pixel area for each light and dark pixel area.

  The determination unit 1206 determines whether or not each gray pixel region is a GGO region based on the representative value of the distance calculated in each gray pixel region. When the determination unit 1206 compares with a predetermined threshold value and determines that the representative value of distance is smaller than the predetermined threshold value, the determination unit 1206 determines that the gray pixel region is not a GGO region (is a normal region). On the other hand, when the determination unit 1206 determines that the representative value of the distance is equal to or greater than the predetermined threshold, the determination unit 1206 determines that the gray pixel region is a GGO region.

  The feature point output unit 1207 determines whether the feature points extracted by the feature point extraction unit 603 are included in the normal region or the GGO region. The feature point output unit 1207 outputs the feature points determined to be included in the normal region to the feature point matching unit 605. Note that the feature point output unit 1207 deletes the feature points determined to be included in the GGO region and does not output them to the feature point matching unit 605. As a result, the feature points determined to be included in the GGO region are excluded from the feature points used for matching in the feature point matching unit 605.

  Next, the flow of the feature point filtering process in the feature point filtering unit 604 will be described. FIG. 13 is a flowchart of the filtering process in the feature point filtering unit. In the flowchart illustrated in FIG. 13, the image extracted by the region extraction unit 602 is input to the feature point filtering unit 604, and the feature point extracted by the feature point extraction unit 603 is input to the feature point filtering unit 604. Processing is started.

  In step S1301, the edge detection unit 1201 performs edge detection processing on the image of the region extracted by the region extraction unit 602, and extracts an edge pixel region. In addition, the edge detection unit 1201 outputs the extracted edge pixel region to the distance calculation unit 1205.

  In step S1302, the density threshold processing unit 1202 performs density threshold processing on the image of the region extracted by the region extraction unit 602, and extracts light and dark pixels.

  In step S1303, the labeling unit 1203 collects the gray pixels extracted in step S1303 for each region to form a gray pixel region, and performs labeling for each gray pixel region.

  In step S <b> 1304, the minute point removing unit 1204 removes the shaded pixel region having a size equal to or smaller than the predetermined size from the labeled shaded pixel region, and outputs the shaded pixel region larger than the predetermined size to the distance calculating unit 1205.

  In step S1305, the distance calculation unit 1205 superimposes the edge pixel region output in step S1301 and the grayscale pixel region output in step S1304. Further, the distance calculation unit 1205 substitutes 1 for the label counter n.

  In step S1306, the distance calculation unit 1205 calculates a representative value of the distance of the edge pixel area included in the gray pixel area from the outline of the gray pixel area for the gray pixel area with the label n = 1. .

  In step S1307, the determination unit 1206 determines whether the representative value calculated in step S1306 is greater than or equal to a predetermined distance threshold. If it is determined in step S1307 that the distance is equal to or greater than the predetermined distance threshold (Yes in step S1307), the process proceeds to step S1308.

  In step S1308, the determination unit 1206 determines that the light and dark pixel region of the label of n = 1 is the GGO region. In step S1309, the feature point output unit 1207 deletes the feature point included in the grayscale pixel area with the label n = 1, and the process advances to step S1312.

  On the other hand, if it is determined in step S1307 that the distance is smaller than the predetermined distance threshold (No in step S1307), the process proceeds to step S1310.

  In step S <b> 1310, the determination unit 1206 determines that the gray pixel area of the label of n = 1 is not the GGO area (determines that it is a normal area). In step S1311, the feature point output unit 1207 outputs the feature point included in the grayscale pixel region with the label of n = 1 to the feature point matching unit 605, and the process proceeds to step S1312.

  In step S1312, the distance calculation unit 1205 determines whether or not the processing in steps S1306 to S1311 has been executed for all the grayscale pixel regions output in step S1304.

  If it is determined in step S1312 that there is a gray pixel region that has not been executed (No in step S1312), the process proceeds to step S1313, the label counter n is incremented, and the process returns to step S1306.

  On the other hand, if it is determined in step S1312 that the process has been executed for all the gray pixel regions (Yes in step S1312), the feature point filtering process is terminated.

  Next, the result of the feature point filtering process by the feature point filtering unit 604 will be described. FIG. 14 is a diagram illustrating images before and after the feature point filtering process. FIG. 14A shows an image after the feature point extraction unit 603 performs the feature point extraction process on the image of the predetermined region 401 extracted by the region extraction unit 602 and before the feature point filtering process. Show. In the case of FIG. 14A, in addition to the feature points extracted in the blood vessels and bronchi in the normal region, the feature points extracted in the blood vessels and bronchi in the GGO region are included.

  On the other hand, FIG. 14B shows an image after the feature point filtering process is performed by the feature point filtering unit 604. In the case of FIG. 14B, feature points extracted in blood vessels and bronchi in the GGO area are deleted.

  As described above, by performing the feature point filtering process, feature points that are not effective for local alignment are deleted, so that the number of feature points used for local alignment can be reduced.

  As is clear from the above description, in the medical image display system 120 according to the first embodiment, the second registration unit 142 includes the feature point filtering unit 604. Further, in the medical image display system 120 according to the first embodiment, the feature point filtering unit 604 is based on the distance from the contour of the gray pixel area of the edge pixel area included in the gray pixel area extracted from the medical image. Thus, it is determined whether or not the light and dark pixel region is a GGO region. Also, the feature point filtering unit 604 removes the feature points included in the gray pixel region determined to be the GGO region as feature points that are not effective for local alignment.

  Thereby, according to the medical image display system 120 in the first embodiment, since feature points that are highly likely to fail in matching are excluded, it is possible to reduce the number of feature points used for local alignment. It becomes possible. As a result, the matching of feature points can be speeded up (that is, local positioning can be speeded up).

  Furthermore, the medical image display system 120 according to the first embodiment uses the feature points included in the light and dark pixel region determined by the feature point filtering unit 604 not to be a GGO region (a normal region). Perform alignment.

  Thereby, according to the medical image display system 120 in 1st Embodiment, it becomes possible to perform local alignment using the feature point effective in alignment. That is, it is possible to eliminate the influence of the temporal change in the local alignment for correcting the position fluctuation accompanying the heartbeat or respiration of the patient. As a result, the accuracy of local alignment can be improved.

[Second Embodiment]
In the first embodiment, the feature points included in the GGO region are excluded from the feature points used for local alignment by deleting the feature points included in the GGO region. On the other hand, in the second embodiment, the reliability is calculated for the feature points included in the GGO region, and is used for local alignment when the calculated reliability satisfies a predetermined condition. . In the second embodiment, the reliability is calculated for each feature point included in the GGO region according to the distance from the contour of the gray pixel region to each feature point. Hereinafter, the second embodiment will be described focusing on the differences from the first embodiment.

  First, the functional configuration of the feature point filtering unit 604 will be described. FIG. 15 is a second diagram illustrating details of the functional configuration of the feature point filtering unit. The difference from the functional configuration of the feature point filtering unit shown in FIG. 12 is that a reliability calculation unit 1501 is provided.

  The reliability calculation unit 1501 outputs, to the feature point output unit 1207, feature points included in the light and shade pixel region that is determined not to be a GGO region (is a normal region) by the determination unit 1206. Further, the reliability calculation unit 1501 outputs the distance from the contour of the grayscale pixel area output from the distance calculation unit 1205 for the feature points included in the grayscale pixel area determined to be the GGO area by the determination unit 1206. Based on the above, the reliability of each feature point is calculated. In addition, the reliability calculation unit 1501 extracts feature points whose feature point reliability is equal to or higher than a predetermined threshold, and outputs the feature points to the feature point output unit 1207 in association with the reliability.

  As a result, the feature point output unit 1207 performs feature point matching between feature points included in the normal region and feature points having reliability equal to or higher than a predetermined threshold among the feature points included in the GGO region. The data can be output to the unit 605.

  As a result, the feature point matching unit 605 can add the feature points included in the GGO region to the feature points used for matching. Further, when the totaling unit 606 calculates the movement vectors of the respective feature points and calculates one movement vector, weighting can be performed according to the respective feature points.

  FIG. 16 is a diagram showing the reliability calculated for each feature point. In FIG. 16, feature points 1601 to 1605 are feature points included in the light and dark pixel region determined to be the GGO region 910.

  The reliability calculation unit 1501 calculates the pixels corresponding to the feature points 1601 to 1605 among the distances from the outlines of the pixels included in the gray pixel region corresponding to the GGO region 910 calculated by the distance calculation unit 1205. Get the distance from the contour. In addition, the reliability calculation unit 1501 calculates the reliability corresponding to the distance from the contour for each of the feature points 1601 to 1605.

  The example of FIG. 16 indicates that the reliability of the feature points 1601, 1602, and 1603 is calculated as 0.6, and the reliability of the feature point 1604 is calculated as 0.1. Further, the example of FIG. 16 shows that the reliability of the feature point 1605 is calculated as 0.7.

  Next, the flow of the feature point filtering process in the second embodiment will be described. FIG. 17 is a second flowchart of the feature point filtering process. The difference from the flowchart shown in FIG. 13 is steps S1701 to S1703.

  In step S <b> 1701, the reliability calculation unit 1501 includes the feature points included in the light and dark pixel area among the distances from the contour of the light and dark pixel area of each pixel included in the light and dark pixel area with the label n = 1. Get the distance for the pixel corresponding to.

  In step S1702, the reliability calculation unit 1501 calculates the reliability of each feature point based on the distance between the feature points acquired in step S1701, and outputs the feature point having the reliability equal to or higher than a predetermined threshold. Output to the unit 1207. The feature point output unit 1207 outputs the reliability calculated by the reliability calculation unit 1501 to the feature point matching unit 605 in association with the feature points.

  In step S <b> 1703, the distance calculation unit 1205 determines whether or not the reliability has been calculated for all feature points included in the grayscale pixel area with the label of n = 1. If it is determined in step S1703 that there is a feature point whose reliability has not been calculated, the process returns to step S1701.

  On the other hand, if it is determined that the reliability has been calculated for all the feature points included in the gray pixel area of the label n = 1, the process proceeds to step S1312.

  As is clear from the above description, in the medical image display system 120 according to the second embodiment, the second registration unit 142 includes the feature point filtering unit 604. In the medical image display system 120 according to the second embodiment, the feature point filtering unit 604 is based on the position of the edge pixel area included in the gray pixel area extracted from the medical image in the gray pixel area. Then, it is determined whether or not the light and dark pixel region is a GGO region. The feature point filtering unit 604 calculates the reliability of the feature point according to the distance from the contour of the gray pixel area to the feature point for the feature point included in the gray pixel area determined to be the GGO area. To do. Furthermore, the feature point filtering unit 604 removes feature points that do not have a predetermined reliability as feature points that are not effective for local alignment.

  Thereby, according to the medical image display system 120 in the second embodiment, since feature points that are highly likely to fail in matching are excluded, it is possible to reduce the number of feature points used for local alignment. It becomes possible. As a result, the matching of feature points can be speeded up (that is, local positioning can be speeded up).

  In addition, the medical image display system 120 according to the second embodiment performs local alignment using the feature points included in the grayscale pixel region that the feature point filtering unit 604 has determined not to be the GGO region. The feature point filtering unit 604 performs local alignment using feature points having a predetermined reliability among feature points included in the gray pixel region determined to be the GGO region. .

  Thereby, according to the medical image display system 120 in the second embodiment, local alignment is performed using feature points that are effective for alignment and feature points that are not effective for alignment but have high reliability. It becomes possible. As a result, the accuracy of local alignment can be improved.

  Furthermore, in the medical image display system 120 according to the second embodiment, the feature points filtering unit 604 adds up the feature points included in the gray pixel region determined to be the GGO region according to the reliability of the feature points. To perform local alignment.

  Thereby, according to the medical image display system 120 in the second embodiment, in the local alignment, it is possible to suppress the influence of the feature points that are not effective for the alignment, thereby improving the accuracy of the local alignment. Can do.

[Other Embodiments]
In the first and second embodiments, the distance calculation unit 1205 calculates the distance from the upper left corner contour to the edge pixel region of the gray pixel region and the distance from the lower right corner contour to the edge pixel region. However, the distance from the contour in another arbitrary direction may be calculated. That is, the determination unit 1206 determines whether or not the gray pixel region is a GGO region by specifying the relative position of the edge pixel region in the gray pixel region.

  In the second embodiment, when the reliability calculation unit 1501 calculates the reliability of the feature points included in the gray pixel area determined to be the GGO area, the characteristic calculation is performed from the contour of the gray pixel area. The distance to the point was used. However, in calculating the reliability, the distance from the center position of the gray pixel area to the feature point may be used. That is, the reliability calculation unit 1501 may calculate the reliability by specifying the relative position of the feature point in the gray pixel region.

  In the second embodiment, the reliability calculation unit 1501 calculates the reliability for the feature points included in the light and dark pixel region determined to be the GGO region. However, the reliability may be calculated in the same manner for the feature points included in the light and dark pixel area determined not to be the GGO area (normal area).

  In the first and second embodiments, the case where a CT image is displayed has been described. However, the present invention may be applied to a case where a medical image other than a CT image (for example, an MRI (Magnetic Resonance Imaging) image) is displayed. Good.

In addition, in the disclosed technology, forms such as the following supplementary notes are conceivable.
(Appendix 1)
A feature point extraction unit for extracting a feature point used for alignment with another medical image from the medical image;
From the medical image, a density threshold processing unit that extracts a gray pixel region including pixels having density values within a predetermined density range;
From the medical image, an edge detection unit that detects an edge pixel region;
Determination that determines whether or not the gray pixel area is a disease area by specifying the position of the edge pixel area included in the gray pixel area extracted from the medical image in the gray pixel area And
Among the feature points extracted from the medical image, the medical image and the other medical image are used by using the feature point included in the gray pixel region determined not to be a disease region by the determination unit. A medical image display system comprising: an alignment processing unit that performs alignment of
(Appendix 2)
The determination unit determines whether or not the gray pixel region is a disease region by specifying a position of the edge pixel region included in the gray pixel region and a position of a contour of the gray pixel region. The medical image display system according to appendix 1.
(Appendix 3)
When the determination unit determines that the distance from the contour of the gray pixel area to the edge pixel area included in the gray pixel area is smaller than a predetermined threshold, the gray pixel area is not a disease area. The medical image display system according to appendix 2, wherein
(Appendix 4)
The image processing apparatus further includes a calculation unit that calculates reliability of the feature point based on a position of the feature point included in the grayscale pixel region determined to be a disease region by the determination unit in the grayscale pixel region. And
In addition to the feature points included in the gray pixel area determined not to be a disease area by the determination unit, the alignment processing unit is determined to be the disease pixel area determined to be a disease area by the determination unit. The medical image display system according to supplementary note 1, wherein the medical image and the other medical image are aligned using the characteristic point included in the image and having a predetermined reliability.
(Appendix 5)
The calculation unit according to claim 4, wherein the calculation unit calculates the reliability of the feature point by specifying the position of the feature point included in the gray pixel region and the position of the contour of the gray pixel region. Medical image display system.
(Appendix 6)
The medical image display system according to supplementary note 5, wherein the calculation unit calculates the reliability of the feature point based on a distance from an outline of the gray pixel region to the feature point included in the gray pixel region. .
(Appendix 7)
Extract feature points used for alignment with other medical images from medical images,
From the medical image, extract a grayscale pixel region including pixels having a density value within a predetermined density range,
An edge pixel region is detected from the medical image,
By determining the position of the edge pixel area included in the gray pixel area extracted from the medical image within the gray pixel area, it is determined whether the gray pixel area is a disease area,
Of the feature points extracted from the medical image, the feature point included in the gray pixel region determined not to be a disease region is used to align the medical image with the other medical image. Do,
A medical image display program for causing a computer to execute processing.
(Appendix 8)
Computer
Extract feature points used for alignment with other medical images from medical images,
From the medical image, extract a grayscale pixel region including pixels having a density value within a predetermined density range,
An edge pixel region is detected from the medical image,
By determining the position of the edge pixel area included in the gray pixel area extracted from the medical image within the gray pixel area, it is determined whether the gray pixel area is a disease area,
Of the feature points extracted from the medical image, the feature point included in the gray pixel region determined not to be a disease region is used to align the medical image with the other medical image. Do,
A medical image display method for executing processing.

  Note that the present invention is not limited to the configurations shown here, such as combinations with other elements, etc., in the configurations described in the above embodiments. These points can be changed without departing from the spirit of the present invention, and can be appropriately determined according to the application form.

100: CT image photographing system 110: CT apparatus 120: Medical image display system 130: Image DB
140: diagnosis support unit 141: first registration unit 142: second registration unit 143: display control unit 300: parallel display screen 401: predetermined region 402: corresponding region 601: CT image input unit 602: region extraction unit 603: Feature point extraction unit 604: Feature point filtering unit 605: Feature point matching unit 606: Total unit 607: Conversion unit 608: Output unit 610: Local alignment processing unit 1201: Edge detection unit 1202: Density threshold processing unit 1203 : Labeling unit 1204: Minute point removal unit 1205: Distance calculation unit 1206: Determination unit 1207: Feature point output unit 1501: Reliability calculation unit

Claims (6)

A feature point extraction unit for extracting a feature point used for alignment with another medical image from the medical image;
From the medical image, a density threshold processing unit that extracts a gray pixel region including pixels having density values within a predetermined density range;
From the medical image, an edge detection unit that detects an edge pixel region;
Determination that determines whether or not the gray pixel area is a disease area by specifying the position of the edge pixel area included in the gray pixel area extracted from the medical image in the gray pixel area And
Among the feature points extracted from the medical image, the medical image and the other medical image are used by using the feature point included in the gray pixel region determined not to be a disease region by the determination unit. A medical image display system comprising: an alignment processing unit that performs alignment of   When the determination unit determines that the distance from the contour of the gray pixel area to the edge pixel area included in the gray pixel area is smaller than a predetermined threshold, the gray pixel area is not a disease area. The medical image display system according to claim 1, which is determined as follows. The image processing apparatus further includes a calculation unit that calculates reliability of the feature point based on a position of the feature point included in the grayscale pixel region determined to be a disease region by the determination unit in the grayscale pixel region. And
In addition to the feature points included in the gray pixel area determined not to be a disease area by the determination unit, the alignment processing unit is determined to be the disease pixel area determined to be a disease area by the determination unit. The medical image display system according to claim 1, wherein the medical image and the other medical image are aligned using the feature point included in the image and having a predetermined reliability.   The medical image display according to claim 3, wherein the calculation unit calculates the reliability of the feature point based on a distance from an outline of the gray pixel region to the feature point included in the gray pixel region. system. Extract feature points used for alignment with other medical images from medical images,
From the medical image, extract a grayscale pixel region including pixels having a density value within a predetermined density range,
An edge pixel region is detected from the medical image,
By determining the position of the edge pixel area included in the gray pixel area extracted from the medical image within the gray pixel area, it is determined whether the gray pixel area is a disease area,
Of the feature points extracted from the medical image, the feature point included in the gray pixel region determined not to be a disease region is used to align the medical image with the other medical image. Do,
A medical image display program for causing a computer to execute processing. Computer
Extract feature points used for alignment with other medical images from medical images,
From the medical image, extract a grayscale pixel region including pixels having a density value within a predetermined density range,
An edge pixel region is detected from the medical image,
By determining the position of the edge pixel area included in the gray pixel area extracted from the medical image within the gray pixel area, it is determined whether the gray pixel area is a disease area,
Of the feature points extracted from the medical image, the feature point included in the gray pixel region determined not to be a disease region is used to align the medical image with the other medical image. Do,
A medical image display method for executing processing.