JP6246287B1 - Medical image processing apparatus and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To efficiently perform medical image processing. A storage unit that stores a medical image having three-dimensional information captured by a medical image capturing device, a cross-sectional image display unit that displays a cross-sectional image of the medical image stored in the storage unit, and a cross-section of the medical image A cross-sectional image of a medical image, and a region image generation unit that detects a luminance, color tone, and contrast of the image and generates a region image having a combination of luminance, color tone, and contrast different from the detected luminance, color tone, and contrast combination A drawing image forming unit that forms a drawing image, a processing command input unit that inputs a processing command for correction processing, and a correction region determination unit that determines a correction region based on the drawing image and the processing command are determined. A correction processing unit that performs correction on the correction area, and the correction processing unit selects a processing command from the deletion process, the addition process, and the additional deletion process for the correction area. The cross-sectional image display unit superimposes and displays the result of the correction process on the cross-sectional image. [Selection] Figure 1

Description

  The present invention relates generally to medical image processing apparatuses and the like, and more specifically to a medical image processing apparatus and the like that performs correction display on a region of interest in medical images.

  In the medical field, the region of interest (organ, lesion, etc.) on the medical image displayed on the display unit of the medical image processing apparatus is automatically extracted by software, and the volume and area of the extracted region are measured. Opportunities for using measurement data as auxiliary information for diagnosis and surgery are increasing. Conventionally, a region of interest automatically extracted often deviates from a region or range that the extracted region is considered to be optimal from the viewpoint of an interpreting doctor. In this case, the region has been manually corrected.

  In addition, when the machine learning function is implemented in the automatic extraction software, when learning the extraction of the attention area, the correct answer area is manually extracted and created in advance as a correct image. It was necessary to learn. In general, many correct images need to be prepared, and manual creation of correct images has been a heavy burden.

  Based on the above-described situation, an image drawing apparatus has been proposed in which a medical image can be drawn, a drawing area can be deleted or added, or the like can be corrected by a simple operation (Patent Document 1).

  That is, Patent Document 1 discloses a display unit that displays an image having a plurality of pixel value regions including a region of interest, and passes through the region of interest and at least one region different from the region of interest on the image. If it is determined that the setting means for setting a simple figure and both end points of the set figure are present in the attention area, the attention area is expanded to the position of the figure, and the set figure If it is determined that both of the two end points do not exist in the region of interest, one of the regions of interest divided by the graphic is deleted, and the display means is changed to the deformed region of interest as a new region of interest. There is disclosed an image drawing apparatus comprising a control means for displaying the image.

Japanese Patent No. 4368378

However, the image drawing apparatus described in Patent Document 1 determines whether to delete or add a region of interest depending on the positions of both end points of the drawn figure. If it is inside, the attention area cannot be deleted. Furthermore, there is a problem that the attention area cannot be added when the start point and the end point of the drawing belong to the background.
In addition, a drawing curve is written along the boundary line of the region of interest, and for example, deletion and addition of a region cannot be performed at the same time, and further improvement of the image processing function has been expected.

  In particular, when organs, bones, blood vessels, and the like are automatically extracted by a computer, it is often the case that region extraction is not optimally performed, and correction thereof must be performed manually. In such correction work, it is desired to be able to work more simply, quickly and efficiently.

  Therefore, a medical image processing apparatus according to an embodiment of the present invention includes a storage unit that stores a medical image having three-dimensional information captured by a medical image capturing apparatus, and a cross section of the medical image stored in the storage unit. A cross-sectional image display unit for displaying an image; and a luminance, color tone, and contrast of the cross-sectional image of the medical image are detected, and a combination of brightness, color tone, and contrast different from the detected luminance, color tone, and contrast combination A region image generation unit that generates a region image, a drawing image formation unit that forms a drawing image on the cross-sectional image of the medical image, a processing command input unit that inputs a processing command for correction processing, and the drawing image And a correction area determination unit that determines a correction area based on the processing command, and a correction processing unit that corrects the determined correction area, The correct processing unit executes a correction process selected by the processing command from the deletion process, the addition process, and the additional deletion process for the correction area, and the cross-sectional image display unit displays the result of the correction process as the cross-section. It is characterized by being displayed superimposed on the image.

  According to a medical image processing apparatus according to an embodiment of the present invention, an intuitive operation such as drawing a handwritten line on a medical image using an input device such as a tablet or a mouse can be performed smoothly and efficiently on a medical image. There is an effect that various processes such as deletion and addition of the attention area can be performed.

It is explanatory drawing explaining the functional block structure of the medical image processing apparatus in one Embodiment of this invention. It is explanatory drawing explaining the mode of the image processing of the medical image processing apparatus in one Embodiment of this invention. It is explanatory drawing explaining the mode of the image processing of the medical image processing apparatus in one Embodiment of this invention. It is explanatory drawing explaining the mode of the other image processing of the medical image processing apparatus in one Embodiment of this invention. It is explanatory drawing explaining the mode of the other image processing of the medical image processing apparatus in one Embodiment of this invention. It is explanatory drawing explaining the mode of the other image processing of the medical image processing apparatus in one Embodiment of this invention. It is explanatory drawing explaining the image processing concept of the medical image processing apparatus in one Embodiment of this invention. It is explanatory drawing explaining the image processing concept of the medical image processing apparatus in one Embodiment of this invention. It is a flowchart explaining the operation | movement flow of the medical image processing apparatus in one Embodiment of this invention. It is a flowchart explaining the operation | movement flow of the medical image processing apparatus in one Embodiment of this invention. It is explanatory drawing explaining the coordinate system and various cross sections which the medical image processing apparatus in one Embodiment of this invention employ | adopts. It is explanatory drawing explaining the example of an image display of the medical image processing apparatus in one Embodiment of this invention. It is explanatory drawing explaining the image processing concept of the medical image processing apparatus in one Embodiment of this invention. It is explanatory drawing explaining the image processing concept of the medical image processing apparatus in one Embodiment of this invention. It is explanatory drawing explaining the image processing concept of the medical image processing apparatus in one Embodiment of this invention. It is explanatory drawing explaining the image processing concept of the medical image processing apparatus in other embodiment of this invention. It is explanatory drawing explaining the mode of the image processing of the medical image processing apparatus in other embodiment of this invention. It is explanatory drawing explaining the image processing concept of the medical image processing apparatus in other embodiment of this invention. It is explanatory drawing explaining the mode of the image processing of the medical image processing apparatus in other embodiment of this invention.

  An embodiment for implementing a medical image processing apparatus according to the present invention will be described in detail with reference to the drawings.

First, FIG. 1 shows a functional block configuration of a medical image processing apparatus according to an embodiment of the present invention.
In FIG. 1, a medical image processing apparatus 100 typically employs a PC (personal computer), a tablet-type terminal device, etc., and a central processing unit 101 such as a CPU, which is a central unit of computer processing, and various hardware. A control unit 102 that executes control, an image display unit (display) 103 such as a display for displaying an image, an SSD that stores image data, a program, and the like, a storage unit 104 such as a hard disk (HDD), and data A main storage unit 105 such as a RAM for temporarily storing data, an instruction unit 106 such as a mouse or pen, an input unit 107 such as a keyboard for instructing image processing as a command (described later), and communication with an external device And a communication I / F 108 for performing the above.
In addition, the apparatus 100 can acquire three-dimensional medical image data from the medical image server 190 via the communication I / F 108 and the network 199. In FIG. 1, the network 199 is described as being connected by wire, but wireless connection by a wireless LAN or the like is also possible.

The program or software necessary for implementing the present invention is typically installed or stored in an SSD or HDD in the storage unit 104 such as a PC or a tablet terminal. All or some of the software modules are read out to the main storage unit 105 and are executed by the CPU 101.
Further, the calculation execution need not always be performed by a central processing unit such as a CPU, and an auxiliary arithmetic unit such as a digital signal processor (DSP) (not shown) can be used.

  When an LCD with a touch panel is adopted as the image display unit (display) 103, the touch input position coordinates on the touch input panel are displayed on the processing system (CPU) of the tablet terminal 15 via an input device interface (not shown). Sent to and processed. When a multi-touch input panel is employed, it can be configured such that a plurality of touch points on the panel can be sensed simultaneously.

  Further, here, in order to facilitate understanding of the present invention, a case where CT slice image data is used as three-dimensional medical image data will be described, but of course the present invention is not limited to a specific data format. Slice image data such as MRI can also be used.

(Image processing technology)
The image processing techniques necessary for the implementation of the present invention are basically based on conventional elemental techniques (automatic area extraction, binarization, multi-leveling, edge detection, area determination, area centroid determination, overlap determination, etc.). The software program is installed in the medical image processing apparatus 100 as described above. The processing operation of the medical image processing apparatus according to the embodiment of the present invention is realized by the individual operations of the hardware described above and the cooperative operation of software that implements the image processing technology.

[First embodiment]
(Example of processing for deleting image area (1))
FIG. 2 shows a state of image processing of the medical image processing apparatus in one embodiment of the present invention. FIG. 2A is a slice cross-sectional image of a CT image, and shows a cross section of a human body torso portion. The horizontal axis of the image is the horizontal direction (shoulder width direction) of the human body, and the vertical direction of the image is the front-rear direction of the human body. In FIG. 2 (B), regions of interest (1) and (2) obtained by applying white mesh processing to regions of interest automatically extracted by software for the cross-sectional image shown in FIG. 2 (A) are shown. It is represented. Since the original medical image has a relatively dark hue, white (or white) color, which is the target color, is automatically displayed as the target region display or manual drawing line for the medical image. Has been adopted. Also, mesh processing is adopted as the region image so that the medical image below the region image can be seen through to some extent. When the cross-sectional image is relatively bright, the region image may be represented by a black (or black) mesh.

More specifically, the region image is set and generated so as to have the luminance of the medical image and the target luminance. In other words, the average value of the luminance of the medical image portion that overlaps the area image is obtained, and when the average value is lower than 128, which is half of the maximum luminance (256) that can be output, white (or white) color is adopted and higher than 128 Sometimes control is performed to adopt a black (or black) color.
Alternatively, the region image is generated by detecting the brightness, color tone, and contrast of the cross-sectional image of the medical image, and having a combination of brightness, color tone, and contrast different from the detected brightness, color tone, and contrast combination. Is done.

  Also, other colors can be adopted as the setting and generation method of the region image. Since the medical image represented in this embodiment is a black and white gray image, contrasting pink or yellow may be adopted. At this time, processing may be performed so that the underlying medical image can be seen through the region image. In addition, when the medical image is a high contrast image such as a binary image of a black and white image, the region image can be an intermediate gray image.

  In the above description, the region image is displayed as a mesh, but other than that, it may be displayed using diagonal lines, dotted lines, dots, or the like. Alternatively, an image in which marks such as “☆”, “*”, and “*” are arranged may be used as the region image.

In FIG. 2B, a drawing image input by handwriting using a tablet with a touch panel (or an input device such as a pen or a mouse) is superimposed and displayed (P1 and P2 in FIG. 2 are endpoints, respectively). This is the line that appears as "drawn line"). In this figure, the drawn image is a line drawn with the intention of cutting an unnecessary area.
Next, in order to cut (delete) this unnecessary area, as an example, a lowercase letter 'c' (Cancel) as a delete command can be input from the keyboard. By this command input, the area of each divided region image is measured, and regions other than the largest region (unnecessary regions) are deleted. FIG. 3 shows an area image after the “unnecessary area” is deleted in this way. In FIG. 3, the area image (2) is determined as an unnecessary area and is deleted.

(Image processing command)
In the above description, the deletion process is executed by inputting the deletion command 'c' from the keyboard. However, the present invention is not limited to this, and other keys can be assigned to the deletion process, It is also possible to assign an arbitrary gesture command in, and to correspond to a select menu on the window screen.

In order to understand the present invention, image processing commands employed in one embodiment of the present invention are shown in the following table (detailed operations of the 'a' command and the 'z' command will be described later).

(Example of processing for deleting image area (2))
FIG. 4 shows another image processing state of the medical image processing apparatus according to the embodiment of the present invention. FIG. 4A shows an image in a case where the drawn image inputted by handwriting input or the like is a closed curve (R1 surrounding the region image (2) in FIG. 4A). When a delete command 'c' is input from the keyboard after drawing the closed curve, the area inside the closed curve becomes a correction target area (delete area) regardless of whether or not the area image is cut, and is deleted. As a result, as shown in FIG. 4B, the region image (2) existing in the closed curve is deleted.

  In addition, the region of interest can be clearly recognized by using a pattern on the medical image having a contrast different from that of the medical image as the region image.

  In addition, since the designated area can be deleted from the handwritten line, the operation method matches the intuitive feeling of human beings, and the deletion process can be executed quickly and comfortably for the operator.

  In addition, in this operation, the deleted area is determined as a cut area of the area image regardless of the position of the drawing end point, so there is no need to pay attention to the position of the drawing start point and end point, and drawing can be performed easily. In addition, in the deletion using the closed curve, the region image can be deleted regardless of whether or not the region image is cut, so that operability easy for the operator is realized.

[Second Embodiment]
(Example of processing for adding image area (1))
FIG. 5 shows another image processing state of the medical image processing apparatus according to the embodiment of the present invention. FIG. 5A shows a state in which handwritten lines are drawn so as to compensate for the missing portion of the region image indicated by the white mesh (in FIG. 5A, Q1 and Q2 are the end points, respectively). Is the line that appears as the "drawn line"). In this figure, the drawn image is a line drawn with the intention of adding a lacking region.
Since the drawing image and the region image form a hole surrounded by the drawing image, when the additional processing command 'a' is input from the keyboard, the hole is filled, and FIG. As shown in FIG. 4, the “added region” portion is additionally processed so as to fill the hole.

(Example of processing for adding image area (2))
FIG. 6 shows another image processing state of the medical image processing apparatus according to the embodiment of the present invention. In FIG. 6A, a closed curve is input by handwriting outside the area image (R2 in FIG. 6A). Then, when the additional process command “a” is input from the keyboard, the area (inside R2) is additionally processed as shown in FIG. 6B). In this example, even if the region to be added is separated from the region image, it can be added to the region image by drawing a closed curve.

  If no hole is formed by the drawing image and the region image and the drawing image is not a closed curve, the drawing image itself can be processed to be added as a correction region. In this case, the drawn image itself is added as a region image.

  By using this function, it is possible to add an area just by drawing a line so as to compensate for a missing part of the area image, so that the image can be corrected with an operation close to human intuition, and the image can be corrected smoothly. There is an effect that can be done.

  In this operation, the additional area is a hole formed by the area image and the drawing image regardless of the position of the drawing end point, so there is no need to pay attention to the position of the drawing start point and end point, and drawing can be done easily. .

  Moreover, since the area can be added using the closed curve regardless of the presence or absence of the area image, the additional work can be smoothly realized for the operator.

[Third embodiment]
(Processing example of adding / deleting multiple image areas simultaneously)
In the present embodiment, a method for easily and simultaneously deleting and adding two or more areas will be described. “Simultaneous” here refers to timing that is considered to be substantially simultaneous by processing in a very short time in terms of processing by the computer.

  FIG. 7A is a diagram in which handwritten lines are drawn on a region image (represented as handwritten lines passing through O1 to O8 with S1 and S2 as end points, respectively). When an automatic extraction process of an organ in a medical image is performed by software processing, an area to be extracted may not be extracted and / or an extra area that should not be extracted may be extracted (that is, The target area cannot be extracted with sufficient accuracy). FIG. 7A shows a situation in which a handwritten line is input on the outer shape of a region (appropriate for representing the target organ) captured by the human eye in such a case.

  Next, when an additional deletion command 'z', which means that addition and deletion are processed simultaneously for a plurality of areas, is input with the keyboard, the area outside the handwritten line is deleted and recessed from the handwritten line. The region (inner region in the figure) is subjected to additional processing, and corrected to a region image having a smooth boundary as shown in FIG. 7B.

  Hereinafter, the details of the correction process will be described in detail. FIG. 8A is a diagram in which handwritten lines are drawn on the area image. When an additional deletion command 'z' key is input here, first, as shown in FIG. 8 (B), the area of each region (divided into six regions) cut by the handwritten line is calculated, and the most Delete everything except large areas. Then, it is determined or certified whether or not a hole is formed by superimposing the region image and the drawn image. This is shown in FIG. Then, by performing filling (addition) processing on the certified holes, as shown in FIG. 8 (D), correction processing is performed into a region image having a smooth boundary similar to FIG. 7 (B).

  In this correction procedure, the protrusion can be deleted and the retraction can be added simply by inputting a line that follows the boundary, so that the area can be corrected in a short time and efficiently.

  The overall processing flow for correction in the embodiment described so far is organized as shown in FIG. That is, in FIG. 9, when the process for correction is started (step S901), a medical image is displayed on the display in step S902, and a predetermined algorithm (or designated command) is displayed in step S903. Thus, when a region of interest in the medical image already exists, the region is formed (automatically extracted) as a region image and displayed. In step S904, a drawing designation of a region to be deleted and / or added is accepted by handwriting lines, and in step S905, a deletion and / or addition correction processing command is received from the keyboard. In step S906, the correction area is determined and determined based on the received command, the correction process is performed in step S907, the corrected area image is displayed on the display in step S908, and the process ends. S909).

  In the processing operation flow shown in FIG. 9, correction processing command reception (step S905), correction region determination processing and determination processing (step S906), and region image correction (step S907) are further described with reference to FIG. Detailed description.

  In FIG. 10, since the drawing designation by the handwritten line has already been accepted at the time when the process is started in step S1001 (step S904), the shape of the drawing image designated for drawing is determined in step S1002. That is, in S1002, it is determined whether the drawn image is a closed curve.

  In step S1002, if the drawn image is not a closed curve (No), the correction area setting method (that is, correction area determination / determination) differs depending on the content of the correction processing command. More specifically, the correction processing command is divided into “delete”, “add”, and “addition deletion (or deletion addition)”. On the other hand, if the drawn image is a closed curve (Yes) in step S1002, the process proceeds to step S1008, and the area in the closed curve is determined to be a correction area.

  If the correction processing command is “delete” (Yes in step S1003), the process proceeds to step S1006 to determine whether the drawn image cuts the region image. If cut (Yes), the process proceeds to step S1009, the area of each cut area is calculated, and the area other than the maximum area is determined as a deletion area. If the area image has not been cut (No in step S1006), the process proceeds to step S1010, and the drawn image itself is determined to be a deletion area.

  When the correction processing command is not “delete” but “add” (Yes in step S1004), the process proceeds to step S1007, where the drawing and the region image are superimposed to form a closed hole, that is, a hole. Judgment is made. If a hole is formed (Yes), the process proceeds to step S1011 and the hole area is determined as an additional area. If no hole is formed (No in step S1007), the process advances to step S1011 to determine that the drawn image itself is an additional area.

  Furthermore, when the correction processing command is not “add” but “addition / deletion” (Yes in step S1005), the process proceeds to step S1013, where the drawing image is superimposed on the region image, and the region image is cut by the drawing image. If so, calculate the area of each cut area, determine that the area other than the maximum area is the deleted area, and add these (these) holes if the drawn image and area image form a hole. Judge as area.

  On the other hand, if the correction processing command is not addition or deletion (No in step S1005), the process proceeds to step S1016 to check whether other processing is to be performed (this is expanded when the command is further expanded). It can be implemented as a process for commands).

  After the correction areas are determined by the above processing flow, the process proceeds to step S1014, and correction processes corresponding to the correction process commands for these correction areas are performed. Specifically, if the correction processing command is “delete”, the correction area or the deletion area is deleted. If the correction processing command is “add”, a correction area or an additional area is added. If the correction processing command is “additional deletion”, the deletion area is deleted and the additional area is added.

  Through the above processing flow, a correction area is formed and an area image is created. This region image is displayed superimposed on the medical image.

  With the addition and deletion of correction processing using this function, it is possible to remove the protrusion of the region image and add the dent by simply tracing the organ boundary of the medical image, so it matches the intuitive sense of human beings and quickly The area can be corrected without a sense of incongruity.

  That is, by using this function, the correction process can be performed more efficiently and faster than the deletion process and the addition process alone.

[Fourth embodiment]
(Batch correction processing for multiple consecutive slice images)
In this embodiment, a method of generating and displaying a three-dimensional medical image from three cross sections of a human body and correcting a plurality of images of these three-dimensional medical images at once will be described.

  In order to observe a three-dimensional medical image in a multifaceted manner, the x-axis of the xyz coordinate axes orthogonal to each other is the horizontal direction of the person (shoulder width direction), the y-axis is the front-back direction of the person (the direction passing through the chest and back), and the z-axis Is often taken in the height direction of the person (the direction passing through the head and soles). At this time, the xy plane is referred to as an axial plane, the xz plane is referred to as a coronal plane, and the yz plane is referred to as a sagittal plane.

  FIG. 11 shows a coordinate system and various cross sections employed by the medical image processing apparatus according to an embodiment of the present invention. An image on the Axial plane is called an Axial image, an image on the Coronal plane is called a Coronal image, and an image on the Sagittal plane is called a Sagittal image. In FIG. 11, each of these cross sections is set so as to pass through three-dimensional coordinates (Px, Py, Pz) (in the same figure, it is set to the left eye on the uppermost face). A cross-sectional image is projected and displayed. Each cross section can be freely moved, and an image in each cross section can be displayed on the display as an axial image, a coronal image, or a sagittal image.

  FIG. 12 shows an image display example of the medical image processing apparatus in one embodiment of the present invention. In FIG. 12, a medical image (a face in the figure) and an area image H (an area with a sand-like pattern including a mouth on the face in the figure) are superimposed on the axial image display area 1201. Is displayed. A one-dot chain line in the vertical direction represents a sagittal plane, and a one-dot chain line in the horizontal direction represents a coronal plane. The images in these cross sections are the Coral image displayed in the Coronal image display area 1202 and the Sagittal image displayed in the Sagittal image display area 1203. A medical image (represented as a cylindrical cross section for convenience of explanation) and a region image H are also superimposed on these images. By displaying these three cross-sections simultaneously, it can be easily and accurately determined whether or not the correction of the region image is corrected correctly over the entire image.

In addition, for example, when correction processing is performed on an axial image displayed in the axial image display area 1201 among the three cross sections, a function of automatically reflecting the correction on other slice images is also provided. If implemented, it is possible to accurately and promptly perform batch correction processing on successive slice images.
Specific examples will be described below.

FIG. 13 shows an image processing concept (batch correction process) of the medical image processing apparatus according to an embodiment of the present invention. In order to delete a part of the area image (for example, around the mouth) in the Axial image of FIG. 12, a drawing image is input to the Axial image by handwriting. That is the closed curve drawn in the area image of FIG.
Next, a correction processing command for the designated region image is required. In one embodiment of the present invention, the meaning of executing batch deletion processing for all slices in the z-axis direction in FIG. Enter the command 'C' (uppercase). In this case, since the handwritten input image is a closed curve, all the inside of the closed curve becomes a correction area, and all the slice images (or appropriate instruction means) read into the memory by displaying this correction area on the screen. All the region images that fall within the closed curve are deleted while being translated in the z-axis direction with respect to a plurality of designated slice images). Each cross section of the region image corrected in this way is displayed on the display in real time.

In order to understand the present invention, image processing commands employed in an embodiment of the present invention are shown in the following table (detailed operations of the “A” command and the “Z” command will be described later).
In the above example, an example of the deletion process is given. However, in the case where correction areas of all slices (or a plurality of designated slices) are additionally processed at once, a key 'A' (uppercase) is assigned. be able to. In addition, in the case where correction areas of all slices (or a plurality of designated slices) are added and deleted at once, a key “Z” (upper case) can be assigned.

  The difference from the previous embodiments is that when the capital letter key is used, the correction processing can be performed collectively for all slices (or a plurality of designated slices).

  In this way, all other slices can be corrected simply by correcting one slice, so that an area at a fixed position on the xy plane of a three-dimensional medical image can be deleted or added. In this case, all slice images can be corrected efficiently. Accordingly, the work time can be shortened and an efficient correction work can be performed.

  In order to grasp a medical image having three-dimensional information three-dimensionally by displaying the medical image in three sections, it is effective to observe from three different directions, and the region such as the organ is clearly recognized. And it is easy to correct the area.

  Furthermore, the axial, coronal, and sagittal cross-sections are the most representative cross-sectional images of medical images, and the normals of the three cross-sections are orthogonal to each other. This has the effect of easily correcting areas such as organs. Further, since the correction result can be observed in real time from different cross-sections via the three-dimensional image, the quality of the correction result can be judged quickly and smoothly, and the correction work can be carried out efficiently.

[Fifth embodiment]
(Other batch correction processing for multiple consecutive slice images)
In the present embodiment, a method for efficiently performing a correction operation using the characteristics of a human body shape or an organ shape will be described with reference to FIGS.

  FIG. 14 shows an image processing concept of the medical image processing apparatus according to the embodiment of the present invention. In FIG. 14, the region images g1 to g4 are displayed in a stacked manner. In general, when a three-dimensional object is viewed in its cross section, the cross-sectional area tends to decrease as the distance from the center increases. The area of the sphere is smaller as the cross section is farther from the center. A rectangular parallelepiped, a triangular pyramid, and the like, except for a specific cross section, tend to have a smaller cross-sectional area as the distance from the center increases. Since human body shapes and organs often have such characteristics, in this embodiment, correction processing is performed using these characteristics.

  In FIG. 14, a region image created on an Axial cross section of a three-dimensional medical image is shown. Next, it is assumed that the correction area is drawn by handwriting on the axial cross section with z coordinate = z1. Next, in one embodiment of the present invention, the Ctrl key and the 'c' key are simultaneously input as commands for performing continuous deletion processing away from the maximum area position (that is, the Ctrl + 'c' key). ). The meaning of this correction processing command is to execute the deletion processing by translating the drawn image in the direction away from the maximum area.

When such a command input is accepted, first, the image position where the region image is maximum in the axial cross section is detected, and the z coordinate of this image is set to zmax. In the drawing, the plane that becomes zmax is located below the Axial cross section with z coordinate = z1. In other words, the Axial cross section with z coordinate = z1 is located on a plane that becomes zmax. This positional relationship determines the direction in which continuous processing (collective processing) described later is to be performed.
That is, in one embodiment of the present invention, the direction of the slide surface that is continuously (collectively) corrected from the cross section (z coordinate = Axial cross section with z1) to which the correction area is input is a plane that is zmax. It is supposed to be away from the direction.

  Next, the same deletion process is performed on each Axial section while sequentially moving the corrected image having the z coordinate z1 away from zmax.

  With such a batch correction process, it is possible to efficiently perform a plurality of deletion processes while reducing the probability of erroneously deleting an area that is not to be deleted. This is because the cross-sectional area located farther from the center is often smaller, so that important portions are not deleted and unnecessary portions are often deleted. In addition, since a small deletion performed at the end portion does not cause a large deletion near the center, there is an effect that an erroneous deletion can be avoided.

  In the above description, an example in which a drawing image is input in an axial section has been described. However, the present invention is not limited to this, and batch correction processing can be similarly performed in a coronal section or a sagittal section. it can. Further, instead of detecting the maximum position of the cross-sectional area, it is also possible to detect the position of the center of gravity of the region image and proceed with the batch correction process in a direction away from the position of the center of gravity (described later with reference to FIG. ).

  As shown in FIG. 15, the correction area is drawn in the Sagittal section whose x coordinate is x1. At this time, the centroid position G of the object (which is an object as a three-dimensional object) is calculated (or calculated in advance), and the Alt key is used as a command for performing continuous deletion processing away from the centroid position G. The 'c' key is input simultaneously (Alt + 'c' key). By this command input, the deletion process is performed while moving the drawing image in a direction away from the center of gravity position G. In FIG. 15, since x1 is located on the left side in the figure from the center of gravity position G, the batch correction process is performed while moving the drawn image from x1 to the left in the figure.

  In this example, the barycentric position of the area image is used. However, the present invention is not limited to this, and the center position may be used. As the center position, a minimum three-dimensional box surrounding the area image can be created, and the coordinates of the midpoint position of each side can be set as the center position.

[Sixth embodiment]
(Further application of batch correction processing)
FIG. 16A shows an embodiment in which a drawing image is input on a plurality of slice images (typically two slice images) of a three-dimensional medical image, and correction processing is performed by interpolating the drawing image between the slices between them. And with reference to FIG. 16B, it demonstrates.

  In this embodiment, first, there is no region image on the medical image, and after inputting a drawing image, the 'a' key is input to add a region. FIG. 16A shows the state after such processing is performed on two slices. That is, in this embodiment, different slice images (not necessarily adjacent slice images) are slices between them so as to connect or fill the specified region image E1 and region image E2 (each a planar region). An attempt is made to apply correction processing to an image.

A more specific embodiment will be described with reference to FIG. 16B.
That is, the upper diagram in FIG. 16B (A) is an image in which the processing for the different slice images described with reference to FIG. 16A is displayed in the sagittal section, and the lower diagram in FIG. 16B (A) is displayed in the coronal section. It is an image. In these sections, the drawn image or the area image in the Axial section is displayed as one line. If one or more Axial cross-sectional slices exist between these lines, the edges of the first two drawn images or region images input to the slice images (their) are infinitely straight lines. A region formed as an intersection (collection) of them when they are connected with each other is interpolated (generated) as a drawing image or a region image of each slice image. When an 's' key indicating inter-slice interpolation (generation) is input, a drawing area between slices is formed and displayed. A state where the inter-slice interpolation is completed is shown in FIG. 16B (B). In FIG. 5B, the area image is displayed as a black mesh. The black color is displayed because the medical image has a bright and whitish tone, and a contrasting black color is automatically adopted (the logic has already been described). In addition, in the sagittal cross section in the upper diagram of FIG. 16B (B) and the coronal cross section in the lower diagram of FIG. 16 (B), an area between the first input two slice images (interpolated for one or more slices) is You can see it being buried.
In the present invention, it is possible to further embed or add to the three-dimensional area specified in this way, which will be further described with reference to FIGS. 17A and 17B.

  FIG. 17A shows a three-dimensional region image V1 created by the procedure described with reference to FIGS. 16A to 16B. A state in which one drawing image or region image E3 is newly added to V1 is schematically shown. As shown in FIG. 16A, when a closed curve is drawn on the slice below the drawn image created in FIG. 16B and an additional processing command 'a' is input, in one embodiment of the present invention, Then, the frustoconical solid region having the bottom surface of the solid region image V1 and the region image E3 as both bottom surfaces is subjected to additional processing.

A more specific embodiment will be described with reference to FIG. 17B.
That is, the upper diagram in FIG. 17B is an image in which the processing described with reference to FIG. 17A is displayed in the Sagittal section, and the lower diagram in FIG. 17B is an image in which the process is displayed in the Coronal section.

  In the state of FIG. 17B (A), at least one between the already created drawing image or region image (a solid composed of a plurality of slice planes) and the newly added drawing image or region image (plane). As a processing command for adding a drawing image to the slice plane, the 's' key is pressed in one embodiment of the present invention. When this 's' key is pressed, a drawing image or region image closest to the newly added drawing image or region image in the already created drawing image or region image (this is a slice because it is a slice) Is selected, and additional processing is performed on one or more slice planes between the drawing image or region image (plane) and the newly added drawing image or region image (plane).

  In the above example, the addition process has been described. However, the present invention is not limited to this, and the deletion process can be similarly performed. In this case, a drawing image input by hand-drawn drawing is accepted on a certain two slice planes, a deletion process by the “c” key as a deletion process command is performed, and then the “s” key is pressed. Thereby, a drawing image or a region image (determined as a deletion region) is inserted between two slices.

  In addition, when an input of a new drawing image or area image is received for the above-described deletion area (solid) and the 's' key is pressed again, a new one of the plurality of slice planes constituting the above-described deletion area is newly created. A drawing image or slice plane closest to the inputted drawing image or area image is detected, and one or more slice planes between the detected slice plane and the new drawing image or area image are further collectively displayed. Deletion processing is performed.

  With the function described above, the cross-sectional area image can be corrected continuously, so that the entire image can be corrected efficiently. Further, for example, by correcting only the section considered important by the interpretation doctor, the section in between is automatically corrected, and the work efficiency is improved.

  In addition, when the correction of the region is started from the vicinity of the center of gravity of the organ toward the end by the above-described function, the region can be corrected continuously using the already corrected image after reaching the end once. In such a case, there is an effect that correction can be made more efficiently.

  These features are mutually exclusive for all of the components described in this specification (including claims, abstract, and drawings) and / or for all steps of all disclosed methods or processes. Except for the combination, any combination can be used.

  Also, each feature described in the specification (including the claims, abstract, and drawings) serves the same purpose, equivalent purpose, or similar purpose, unless expressly denied. Alternative features can be substituted. Thus, unless expressly denied, each feature disclosed is one example only of a generic series of identical or equivalent features.

  Furthermore, the present invention is not limited to any specific configuration of the above-described embodiment. The invention includes all novel features or combinations thereof described in the specification (including claims, abstract, and drawings), or all novel methods or process steps described, or Can be extended to any combination.

101 Central processing unit (CPU)
102 control unit 103 image display unit (display, etc.)
104 Storage unit (SSD, hard disk, ROM, etc.)
105 Main memory (RAM, etc.)
106 Instruction section (mouse, pen, etc.)
107 Input unit (keyboard, etc.)
108 Communication I / F
109 Connection bus or connection line 190 Medical image server 199 Network line

Claims (16)

A storage unit for storing a medical image having three-dimensional information captured by the medical image capturing apparatus;
A cross-sectional image display unit that displays a cross-sectional image of the medical image stored in the storage unit;
A region image generation unit that detects the luminance, color tone, and contrast of the cross-sectional image of the medical image and generates a region image having a combination of luminance, color tone, and contrast that is different from the detected combination of luminance, color tone, and contrast When,
A drawing image forming unit that forms a drawing image on the cross-sectional image of the medical image;
A processing command input section for inputting a processing command for correction processing;
A correction area determination unit that determines a correction area based on the drawing image and the processing command;
A correction processing unit for correcting the determined correction area,
The correction processing unit executes correction processing selected by the processing command from among (1) deletion processing, (2) addition processing, and (3) additional deletion processing for the correction area,
The medical image processing apparatus, wherein the cross-sectional image display unit displays the result of the correction process superimposed on the cross-sectional image. The medical image processing apparatus according to claim 1,
If the processing command is a deletion processing command, the correction area determination unit
(A) When the drawn image forms a closed curve, the entire area inside the closed curve is determined as a correction area,
(B) If the drawn image does not form a closed curve and cuts the region image, determine a region other than the largest region of the cut region image as a correction region,
(C) In cases other than (A) and (B) above, the drawn image itself is determined as a correction area,
The medical image processing apparatus, wherein the correction processing unit executes a deletion process on the determined correction area. The medical image processing apparatus according to claim 1,
If the processing command is an additional processing command, the correction area determination unit
(D) When the drawn image forms a closed curve, the entire area inside the closed curve is determined as a correction area,
(E) When the superimposed image of the drawn image and the area image forms a hole area, the hole area is determined as a correction area;
(F) In cases other than the above (D) and (E), the drawing image itself is determined as a correction area,
The medical image processing apparatus, wherein the correction processing unit executes an additional process for the determined correction area. The medical image processing apparatus according to claim 1,
If the processing command is an additional deletion processing command, the correction area determination unit
(G) When the drawn image cuts the area image, the area other than the largest area of the area image is determined as the correction area,
(H) When the superimposed image of the drawn image and the area image forms a hole area, the hole area is determined as a correction area;
The correction processing unit performs a deletion process on the correction area determined in (G) and performs an additional process on the correction area determined in (H). Medical image processing apparatus. The medical image processing apparatus according to claim 1,
The cross-sectional image display unit displays cross-sectional images on three planes having three normal lines intersecting at one point in a three-dimensional coordinate system of three-dimensional information included in the medical image, respectively. Image processing device.   6. The medical image processing apparatus according to claim 5, wherein the three plane cross-sectional images are an axial image, a sagittal image, and a coronal image in a medical image, respectively. The medical image processing apparatus according to claim 1,
The correction area determination unit determines the correction area with respect to a plurality of cross-sectional area images parallel to the cross-section on which the drawn image is formed;
The medical image processing apparatus, wherein the correction processing unit executes a correction process on the determined correction area. The medical image processing apparatus according to claim 1,
The correction area determination unit continuously moves in a direction away from a specific position of a plurality of cross-sectional area images parallel to the cross-section on which the drawing image is formed, and determines the correction area in the image cross-section during the movement And
The medical image processing apparatus, wherein the correction processing unit executes a correction process on the determined correction area. The medical image processing apparatus according to claim 8, wherein
The medical image processing apparatus according to claim 1, wherein the specific position is a position having a maximum area of a size of a region image in a plurality of cross sections parallel to the cross section on which the drawing image is formed. The medical image processing apparatus according to claim 8, wherein
The medical image processing apparatus, wherein the specific position is a center of gravity position of a medical image having three-dimensional information. The medical image processing apparatus according to claim 1,
When a drawing image is formed in a plurality of the image slices and each drawing image has one or more image slices between adjacent drawing images, the drawing image is interpolated on the one or more image slices. A medical image processing apparatus that forms a corrected image and executes a deletion process or an addition process on the corrected image in accordance with the processing command. In the medical image processing apparatus according to claim 11, after performing a deletion process or an addition process on the corrected image,
Further, when a drawn image is formed in the image cross section and one or more image cross sections exist between the drawn image and the already formed region image, the drawn image is interpolated on the one or more image cross sections. A medical image processing apparatus which forms a corrected image and executes a deletion process or an addition process on the corrected image in accordance with the processing command. A program executed on a computer that performs correction processing on a medical image having three-dimensional information imaged by a medical image imaging device, and when executed by the computer,
Storing a medical image having three-dimensional information captured by the medical image capturing apparatus in a storage unit;
Displaying a cross-sectional image of the medical image stored in the storage unit on a cross-sectional image display unit;
A region image having a combination of brightness, color tone, and contrast different from the detected combination of brightness, color tone, and contrast is detected in the region image generation module by detecting the brightness, color tone, and contrast of the cross-sectional image of the medical image. Generating step;
Causing the drawing image forming module to form a drawing image on the cross-sectional image of the medical image;
Causing the processing command input unit to input a processing command for the correction processing;
Causing a correction area determination unit to determine a correction area based on the drawing image and the processing command;
And a step of causing the correction processing unit to perform correction on the determined correction area,
Causing the correction processing unit to execute correction processing selected by the processing command from among (1) deletion processing, (2) addition processing, and (3) additional deletion processing for the correction area;
A program that causes the cross-sectional image display unit to display a result of the correction process superimposed on the cross-sectional image. If the processing command is a deletion processing command, the correction area determination unit,
(A) When the drawn image forms a closed curve, the entire area inside the closed curve is determined as a correction area,
(B) When the drawing image does not form a closed curve and cuts the region image, the region other than the largest region of the cut region image is determined as the correction region,
(C) In the case other than the above (A) and (B), the step of determining the drawn image itself as a correction region;
The program according to claim 13, further comprising: causing the correction processing unit to execute a deletion process on the determined correction area. If the processing command is an additional processing command, the correction area determination unit
(D) When the drawn image forms a closed curve, the entire area inside the closed curve is determined as the correction area,
(E) When the superimposed image of the drawn image and the area image forms a hole area, the hole area is determined as a correction area;
(F) In the case other than the above (D) and (E), the step of determining the drawn image itself as a correction region;
The program according to claim 13, further comprising: causing the correction processing unit to perform an additional process on the determined correction area. If the processing command is an additional deletion processing command, the correction area determination unit,
(G) When the drawn image cuts the region image, the region other than the largest region of the region image is determined as the correction region,
(H) When the superimposed image of the drawn image and the area image forms a hole area, the hole area is determined as a correction area;
Causing the correction processing unit to execute a deletion process for the correction area determined in (G) and performing an additional process for the correction area determined in (H). The program according to claim 13.