A method for extracting and processing PET / CT and radiotherapy dose images of a lung tumor target region

Abstract

The embodiment of the present application relates to the technical field of medical image processing, in particular to a lung tumor target area PET / CT, radiotherapy dose image extraction and processing method, the method comprises the following steps: extracting the tumor target area image from the whole lung CT image before radiotherapy; registering each pixel point in the tumor target area image with the pixel point in the radiotherapy dose image; registering each pixel point in the tumor target area image with the pixel point in the whole lung PET image before radiotherapy; and registering each pixel point in the tumor target area image with the pixel point in the whole lung CT image during radiotherapy. The method provided by the present application can extract the tumor target area from the whole lung CT image, and convert the CT image before radiotherapy, the CT image during radiotherapy, the PET image before radiotherapy and the radiotherapy dose image data taken by different devices to the same coordinate system, thereby providing image data basis for subsequent tumor radiotherapy reaction research.

Description

Technical Field

[0001] The present application relates to the field of medical image processing technology, and in particular to a method for extracting and processing lung tumor target area PET / CT and radiotherapy dose images. Background Art

[0002] For patients with lung tumors undergoing radiotherapy, multiple PET / CT images of the entire lung are usually taken and the radiotherapy dose is set during the treatment process. However, due to the differences in the coordinate system and slice plane of the CT image taken each time, when implementing tasks such as radiotherapy response prediction, it is often encountered that the pre-radiotherapy and mid-radiotherapy data, PET / CT data, and radiotherapy dose data cannot be matched one to one. In addition, the research object of tasks such as radiotherapy response prediction is usually only the tumor target area, which needs to be extracted from the entire lung image. We designed a method to solve this problem, extracting the tumor target area from the entire lung CT image, and converting the pre-radiotherapy CT image, mid-radiotherapy CT image, pre-radiotherapy PET image, and radiotherapy dose image data taken by different devices to the same coordinate system, providing a data basis for subsequent tumor radiotherapy response research. Summary of the Invention

[0003] The present application provides a lung tumor target area PET / CT and radiotherapy dose image extraction and processing method, which extracts the tumor target area from the entire lung CT image, and converts the pre-radiotherapy CT image, mid-radiotherapy CT image, pre-radiotherapy PET image, and radiotherapy dose image data taken by different equipment into the same coordinate system, providing a data basis for subsequent tumor radiotherapy response research.

[0004] To solve the above technical problems, an embodiment of the present application provides a lung tumor target area PET / CT and radiotherapy dose image extraction and processing method, comprising the following steps: first, extracting the tumor target area image from the entire lung CT image in the early stage of radiotherapy; then, aligning each pixel point in the tumor target area image with the pixel point in the radiotherapy dose image; next, aligning each pixel point in the tumor target area image with the pixel point in the entire lung PET image in the early stage of radiotherapy; finally, aligning each pixel point in the tumor target area image with the pixel point in the entire lung CT in the middle stage of radiotherapy.

[0005] In some exemplary embodiments, a tumor target area image is extracted from an entire lung CT image before radiotherapy, including: outlining the target area portion with discontinuous points in the pre-radiotherapy CT image, and storing the original coordinates of the points in the pre-radiotherapy CT device coordinate system, the origin of the coordinate system, and the actual length per unit pixel; converting the original coordinates of the contour points of the outlined target area portion into image coordinates; filling the contour of the target area portion to obtain the image coordinates of all points inside the entire target area; and converting the image coordinates of all points in the entire target area back to the original coordinates in the pre-radiotherapy CT device coordinate system.

[0006] In some exemplary embodiments, converting the original coordinates of the outline points of the target area into image coordinates includes calculating the position of each point in a 512*512 CT image based on the original coordinates of the point in the coordinate system of the pre-radiotherapy CT device. The corresponding relationship between the coordinates and the position is:

[0007] XY_img=(XY_ct1–Origin_ct1) / spacing_ct1

[0008] Among them, XY_img is the coordinate of the point in the CT image, XY_ct1 is the original coordinate of the point in the coordinate system of the CT device before radiotherapy, Origin_ct1 is the origin of the coordinate system of the CT device before radiotherapy, and spacing_ct1 is the actual length of the unit pixel point in the coordinate system of the CT device before radiotherapy.

[0009] In some exemplary embodiments, filling the outline of the target area portion to obtain the image coordinates of all points inside the entire target area includes: using the OpenCV library to fill the target area outline image to obtain the image coordinates of the points inside the complete target area.

[0010] In some exemplary embodiments, the image coordinates of all points in the entire target area are converted back to the original coordinates in the pre-radiotherapy CT device coordinate system, including: calculating the original coordinates of each point in the pre-radiotherapy CT device coordinate system according to the position of each point in the 512*512 CT image, and the corresponding relationship between the coordinates and the position is:

[0011] XY_ct1=XY_img*spacing_ct1+Origin_ct1

[0012] Among them, XY_img is the coordinate of the point in the CT image, XY_ct1 is the original coordinate of the point in the coordinate system of the CT device before radiotherapy, Origin_ct1 is the origin of the coordinate system of the CT device before radiotherapy, and spacing_ct1 is the actual length of the unit pixel point in the coordinate system of the CT device before radiotherapy.

[0013] In some exemplary embodiments, each pixel point in the tumor target image is aligned with a pixel point in the radiotherapy dose image, including: reading the coordinate values ​​of the target portion in the radiotherapy dose image and the radiotherapy dose value; for each point in the tumor target image, determining the smallest cuboid containing it, and performing weighted trilinear interpolation on the radiotherapy dose value of the target object based on its vertex data, with the weight being the inverse of the distance between the target object and the vertex.

[0014] In some exemplary embodiments, each pixel point in the tumor target area image is registered with the pixel point in the PET image of the entire lung before radiotherapy, including: reading the coordinate value of the target area portion in the PET image; for each point in the tumor target area image, converting the point coordinates in the CT coordinate system before radiotherapy into the point coordinates in the PET coordinate system through the registration matrix; calculating the distance between each point in the target area and each point in the PET image, and selecting the PET value of the nearest point as the PET value corresponding to the target area point.

[0015] In some exemplary embodiments, for each point in the tumor target area image, the point coordinates in the pre-radiotherapy CT coordinate system are converted into point coordinates in the PET coordinate system through a registration matrix, including: forming the point coordinates in the pre-radiotherapy CT coordinate system into a vector (X, Y, Z), multiplying it with the registration matrix to realize the rotation and translation of the coordinate system, and converting it into the point coordinates in the PET coordinate system.

[0016] In some exemplary embodiments, each pixel point in the tumor target area image is aligned with the pixel point in the entire lung CT at the mid-stage of radiotherapy, including: reading the coordinate values ​​of the mid-stage of radiotherapy CT image points and the CT pixel values; for each point in the tumor target area image, converting the point coordinates in the early stage of radiotherapy CT coordinate system into the point coordinates in the mid-stage of radiotherapy CT coordinate system through the alignment matrix; for each point in the tumor target area image, determining the smallest cuboid containing it, and performing weighted trilinear interpolation on the mid-stage of radiotherapy CT pixel values ​​of the target object according to its vertex data, with the weight being the inverse of the distance between the target object and the vertex.

[0017] In some exemplary embodiments, if there is no valid data for the eight vertices of the smallest enclosing cuboid of the target object on the grid, all grid points on the six faces of the next smallest enclosing cuboid are selected as the source of fitting data. If the next smallest enclosing cuboid still cannot achieve the fitting task, a larger enclosing cuboid is selected, and so on, until the fitting task is achieved.

[0018] The technical solution provided by the embodiments of the present application has at least the following advantages:

[0019] The application provides a lung tumor target area PET / CT and radiotherapy dose image extraction and processing method, which comprises the following steps: first, extracting a tumor target area image from a whole lung CT image before radiotherapy; then, registering each pixel point in the tumor target area image with a pixel point in a radiotherapy dose image; next, registering each pixel point in the tumor target area image with a pixel point in a whole lung PET image before radiotherapy; and finally, registering each pixel point in the tumor target area image with a pixel point in a whole lung CT image during radiotherapy. The method provided by the application extracts a tumor target area part from a whole lung CT image, and converts radiotherapy before CT image, radiotherapy during CT image, radiotherapy before PET image and radiotherapy dose image data taken by different devices to the same coordinate system, thereby providing a data basis for subsequent tumor radiotherapy reaction research. BRIEF DESCRIPTION OF DRAWINGS

[0020] One or more embodiments are illustrated by way of example in the figures that form a part of this disclosure and which are shown by way of illustration in the drawings and are not necessarily drawn to scale unless otherwise specifically noted. The figures in the drawings of which are like numbers refer to like components unless otherwise specified.

[0021] Figure 1 A flowchart of a lung tumor target area PET / CT and radiotherapy dose image extraction and processing method provided by an embodiment of the application is shown in the figure.

[0022] Figure 2 A tumor target area CT image extraction and filling schematic diagram provided by an embodiment of the application is shown in the figure.

[0023] Figure 3 A tumor target area PET image schematic diagram provided by an embodiment of the application is shown in the figure.

[0024] Figure 4 A tumor target area radiotherapy dose image schematic diagram provided by an embodiment of the application is shown in the figure. DETAILED DESCRIPTION

[0025] As known from the background, the radiotherapy before CT image, radiotherapy during CT image, radiotherapy before PET image and radiotherapy dose image data are taken by different devices, and when realizing tasks such as radiotherapy reaction effect prediction, the radiotherapy before data and radiotherapy during data, PET / CT data and radiotherapy dose data cannot be one-to-one corresponding, which is a difficult problem. In addition, the research object of the radiotherapy reaction effect prediction task is usually only the tumor target area, which needs to be extracted from the whole lung image.

[0026] In response to the shortcomings of the prior art, an embodiment of the present application provides a lung tumor target area PET / CT and radiotherapy dose image extraction and processing method, comprising the following steps: first, extracting the tumor target area image from the entire lung CT image in the early stage of radiotherapy; then, aligning each pixel in the tumor target area image with the pixel in the radiotherapy dose image; next, aligning each pixel in the tumor target area image with the pixel in the entire lung PET image in the early stage of radiotherapy; finally, aligning each pixel in the tumor target area image with the pixel in the entire lung CT in the mid-stage of radiotherapy. The purpose of the present invention is to provide a lung tumor target area PET / CT and radiotherapy dose image extraction and processing method, extracting the tumor target area portion from the entire lung CT image, and converting the early stage of radiotherapy CT image, mid-stage of radiotherapy CT image, early stage of radiotherapy PET image, and radiotherapy dose image data taken by different devices into the same coordinate system, providing a data basis for subsequent tumor radiotherapy response research.

[0027] The following detailed description of the various embodiments of the present application is provided in conjunction with the accompanying drawings. However, those skilled in the art will appreciate that many technical details are provided in the various embodiments of the present application to facilitate a better understanding of the present application. However, even without these technical details and the various variations and modifications based on the following embodiments, the technical solutions claimed in the present application can still be implemented.

[0028] See Figure 1 The present invention provides a lung tumor target area PET / CT and radiotherapy dose image extraction and processing method, comprising the following steps:

[0029] Step S1: extracting a tumor target area image from the entire lung CT image in the early stage of radiotherapy.

[0030] Step S2: align each pixel point in the tumor target image with the pixel point in the radiotherapy dose image.

[0031] Step S3: align each pixel in the tumor target image with the pixel in the entire lung PET image in the early stage of radiotherapy.

[0032] Step S4: align each pixel point in the tumor target image with the pixel point in the entire lung CT scan during the middle period of radiotherapy.

[0033] Specifically, in step S1, the tumor target area image is extracted from the entire lung CT image before radiotherapy, including the following steps: outlining the target area with discontinuous points in the CT image before radiotherapy, and storing the original coordinates of the points in the coordinate system of the CT device before radiotherapy, the origin of the coordinate system, and the actual length of the unit pixel; converting the original coordinates of the contour points of the outlined target area into image coordinates; filling the contour of the target area to obtain the image coordinates of all points inside the entire target area; and converting the image coordinates of all points in the entire target area back to the original coordinates in the coordinate system of the CT device before radiotherapy.

[0034] It should be noted that the doctor outlines the target area from the entire lung CT image in the early stage of radiotherapy, and converts the original coordinates of the outline points of the target area into image coordinates, including: calculating the position of each point in the 512*512 CT image based on the original coordinates of the CT device in the early stage of radiotherapy. The corresponding relationship between coordinates and positions is:

[0035] XY_img=(XY_ct1–Origin_ct1) / spacing_ct1

[0036] Among them, XY_img is the coordinate of the point in the CT image, XY_ct1 is the original coordinate of the point in the coordinate system of the CT device before radiotherapy, Origin_ct1 is the origin of the coordinate system of the CT device before radiotherapy, and spacing_ct1 is the actual length of the unit pixel point in the coordinate system of the CT device before radiotherapy.

[0037] It should be noted that the outline of the target area is filled to obtain the image coordinates of all points inside the entire target area, including: using the OpenCV library to fill the target area outline image to obtain the image coordinates of the points inside the complete target area.

[0038] It should be noted that the image coordinates of all points in the entire target area are converted back to the original coordinates in the pre-radiotherapy CT device coordinate system, including: according to the position of each point in the 512*512 CT image, its original coordinates in the pre-radiotherapy CT device coordinate system are calculated. The corresponding relationship between coordinates and positions is:

[0039] XY_ct1=XY_img*spacing_ct1+Origin_ct1

[0040] Among them, XY_img is the coordinate of the point in the CT image, XY_ct1 is the original coordinate of the point in the coordinate system of the CT device before radiotherapy, Origin_ct1 is the origin of the coordinate system of the CT device before radiotherapy, and spacing_ct1 is the actual length of the unit pixel point in the coordinate system of the CT device before radiotherapy.

[0041] Specifically, in step S2, each pixel point in the tumor target area image (pre-radiotherapy tumor target area CT image) is registered with the pixel point in the radiotherapy dose image, including: reading the coordinate value and radiotherapy dose value of the target area part in the radiotherapy dose image; for each point in the tumor target area image, determining the smallest cuboid containing it, and performing weighted trilinear interpolation on the radiotherapy dose value of the target object according to the vertex data, and the weight is taken as the reciprocal of the distance between the target object and the vertex.

[0042] It should be noted that for the data missing problem, there may be a case that the target object has no effective data on the eight vertices of the smallest enclosing cuboid on the grid. For such a case, consider the larger enclosing cuboid of the next layer, select all grid points on the six faces of the smaller enclosing cuboid as the source of fitting data. If the smaller enclosing cuboid still cannot achieve the fitting task, take the larger enclosing cuboid, and so on, until the fitting task is achieved.

[0043] Specifically, in step S3, each pixel point in the tumor target area image is registered with the pixel point in the whole lung PET image before radiotherapy, including: reading the coordinate value of the target area part in the PET image; for each point in the tumor target area image, converting the point coordinates in the CT coordinate system before radiotherapy into the point coordinates in the PET coordinate system through the registration matrix; calculating the distance between each point in the target area and each point in the PET image, and selecting the PET value of the nearest point as the PET value corresponding to the target area point.

[0044] It should be noted that for each point in the CT target area image before radiotherapy, the point coordinates in the CT coordinate system before radiotherapy are converted into the point coordinates in the PET coordinate system through the registration matrix, including: forming a vector (X, Y, Z) of the point coordinates in the CT coordinate system before radiotherapy, multiplying it by the registration matrix to realize the rotation and translation of the coordinate system, and converting it into the point coordinates in the PET coordinate system.

[0045] Specifically, in step S4, each pixel point in the tumor target area image is registered with the pixel point in the whole lung CT during radiotherapy, including: reading the point coordinate value and CT pixel value of the radiotherapy CT image; for each point in the CT target area image before radiotherapy, converting the point coordinates in the CT coordinate system before radiotherapy into the point coordinates in the CT coordinate system during radiotherapy through the registration matrix; for each point in the CT target area image before radiotherapy, determining the smallest cuboid containing it, and performing weighted trilinear interpolation on the radiotherapy CT pixel value of the target object according to the vertex data, and the weight is taken as the reciprocal of the distance between the target object and the vertex.

[0046] It should be noted that for the data missing problem, there may be a situation where there is no valid data for the eight vertices of the smallest enclosing cuboid of the target object on the grid. In this case, consider its larger enclosing cuboid and select all the grid points on the six faces of the smaller enclosing cuboid as the source of fitting data. If the smaller enclosing cuboid still cannot achieve the fitting task, then take the larger enclosing cuboid, and so on, until the fitting task is achieved.

[0047] The lung tumor target area PET / CT, radiotherapy dose image extraction, and processing method provided by this application are described in detail below with respect to a specific embodiment.

[0048] First, the lung tumor target area PET / CT, radiotherapy dose image extraction, and processing methods proposed in the embodiments of this application are based on the Windows 10 system, implemented using Python 3.9, and accelerated using the CUDA API.

[0049] In addition, it should be noted that the tumor image data in the embodiment of the present application are PET / CT and radiotherapy dose images before and during radiotherapy of non-small cell lung cancer.

[0050] First, the tumor target area image is extracted from the whole lung CT image in the early stage of radiotherapy. The tumor target area image is formed by the doctor's outline. Figure 2 .

[0051] Then, each pixel in the tumor target area image (CT image of the tumor target area before radiotherapy) is registered with the pixel in the radiotherapy dose image.

[0052] Next, each pixel in the CT image of the tumor target area before radiotherapy is aligned with the pixel in the PET image of the entire lung before radiotherapy.

[0053] Finally, each pixel in the CT image of the tumor target area in the early stage of radiotherapy is aligned with the pixel in the entire lung CT image in the middle stage of radiotherapy.

[0054] See Figure 3 and Figure 4 , Figure 3 A schematic diagram of a PET image of a tumor target area provided in one embodiment of the present application. Figure 4 This is a schematic diagram of a tumor target area radiotherapy dose image provided by an embodiment of the present application. Figure 2 、 Figure 3 and Figure 4As can be seen in the figure, the lung tumor target PET / CT and radiotherapy dose image extraction and processing method provided in this application extracts the tumor target area from the entire lung CT image and converts the pre-radiotherapy CT image, mid-radiotherapy CT image, pre-radiotherapy PET image, and radiotherapy dose image data captured by different equipment into the same coordinate system, providing a data foundation for subsequent tumor radiotherapy response research. Furthermore, the method of this application has been tested, demonstrating its effectiveness.

[0055] Based on the above technical solution, the embodiment of the present application provides a method for extracting and processing lung tumor target area PET / CT and radiotherapy dose images, which includes the following steps: first, extracting the tumor target area image from the entire lung CT image in the early stage of radiotherapy; then, aligning each pixel in the tumor target area image with the pixel in the radiotherapy dose image; next, aligning each pixel in the tumor target area image with the pixel in the entire lung PET image in the early stage of radiotherapy; finally, aligning each pixel in the tumor target area image with the pixel in the entire lung CT in the mid-stage of radiotherapy. The method provided in the present application extracts the tumor target area portion from the entire lung CT image, and converts the pre-radiotherapy CT image, mid-radiotherapy CT image, pre-radiotherapy PET image, and radiotherapy dose image data taken by different devices into the same coordinate system, providing a data basis for subsequent tumor radiotherapy response research.

[0056] Those skilled in the art will appreciate that the above-described embodiments are specific examples for implementing the present application, and that in actual applications, various changes in form and detail may be made thereto without departing from the spirit and scope of the present application. Any person skilled in the art may make changes and modifications without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application shall be subject to the scope defined in the claims.

Claims

1. A method for extracting and processing lung tumor target area PET / CT and radiotherapy dose images, characterized in that: The following steps are involved: Extract tumor target images from the entire lung CT images during the early stages of radiotherapy; registering each pixel point in the tumor target image with the pixel point in the radiotherapy dose image; Registering each pixel in the tumor target image with the pixel in the entire lung PET image before radiotherapy; Registering each pixel in the tumor target image with the pixel in the entire lung CT scan during the middle period of radiotherapy; Extract tumor target images from the entire lung CT images before radiotherapy, including: The target area is outlined with discontinuous points in the pre-radiotherapy CT image, and the original coordinates of the points in the pre-radiotherapy CT device coordinate system, the origin of the coordinate system, and the actual length of the unit pixel are stored; Convert the original coordinates of the outline points of the target area into image coordinates; Fill the outline of the target area to obtain the image coordinates of all points inside the entire target area; Convert the image coordinates of all points in the entire target area back to the original coordinates in the CT device coordinate system before radiotherapy; Registering each pixel point in the tumor target image with a pixel point in the radiotherapy dose image includes: Read the coordinate values ​​of the target area and the radiotherapy dose value in the radiotherapy dose image; For each point in the tumor target image, determine the smallest cuboid containing it, and perform weighted trilinear interpolation on the radiotherapy dose value of the target object based on its vertex data, with the weight being the inverse of the distance between the target object and the vertex; Registering each pixel in the tumor target image with the pixel in the entire lung PET image before radiotherapy, including: Read the coordinate values ​​of the target area in the PET image; For each point in the tumor target image, the point coordinates in the CT coordinate system before radiotherapy are converted into the point coordinates in the PET coordinate system through the registration matrix; Calculate the distance between each point in the target area and each point in the PET image, and select the PET value of the closest point as the PET value corresponding to the target point; Registering each pixel in the tumor target image with the pixel in the entire lung CT scan during the interim radiotherapy period includes: Read the coordinate values ​​of CT image points and CT pixel values ​​during the middle stage of radiotherapy; For each point in the tumor target image, the point coordinates in the early radiotherapy CT coordinate system are converted into the point coordinates in the mid-radiotherapy CT coordinate system through the registration matrix; For each point in the tumor target image, the smallest cuboid containing it is determined, and the mid-term radiotherapy CT pixel value of the target object is subjected to weighted trilinear interpolation according to its vertex data, with the weight being the inverse of the distance between the target object and the vertex.

2. The lung tumor target area PET / CT and radiotherapy dose image extraction and processing method according to claim 1, characterized in that: Convert the original coordinates of the outline points of the target area into image coordinates, including: According to the original coordinates of each point in the coordinate system of the CT device before radiotherapy, its position in the 512*512 CT image is calculated. The corresponding relationship between coordinates and positions is: XY_img = (XY_ct1 – Origin_ct1) / spacing_ct1 Among them, XY_img is the coordinate of the point in the CT image, XY_ct1 is the original coordinate of the point in the coordinate system of the CT device before radiotherapy, Origin_ct1 is the origin of the coordinate system of the CT device before radiotherapy, and spacing_ct1 is the actual length of the unit pixel point in the coordinate system of the CT device before radiotherapy.

3. The lung tumor target area PET / CT and radiotherapy dose image extraction and processing method according to claim 1, characterized in that: Fill the outline of the target area to obtain the image coordinates of all points inside the entire target area, including: The OpenCV library is used to fill the target area contour image and obtain the image coordinates of the internal points of the complete target area.

4. The lung tumor target area PET / CT and radiotherapy dose image extraction and processing method according to claim 1, characterized in that: Convert the image coordinates of all points in the entire target area back to the original coordinates in the CT device coordinate system before radiotherapy, including: According to the position of each point in the 512*512 CT image, its original coordinates in the coordinate system of the CT device before radiotherapy are calculated. The corresponding relationship between coordinates and positions is: XY_ct1 = XY_img *spacing_ct1 +Origin_ct1 Among them, XY_img is the coordinate of the point in the CT image, XY_ct1 is the original coordinate of the point in the coordinate system of the CT device before radiotherapy, Origin_ct1 is the origin of the coordinate system of the CT device before radiotherapy, and spacing_ct1 is the actual length of the unit pixel point in the coordinate system of the CT device before radiotherapy.

5. The lung tumor target area PET / CT and radiotherapy dose image extraction and processing method according to claim 1, characterized in that: For each point in the tumor target image, the point coordinates in the pre-radiotherapy CT coordinate system are converted into the point coordinates in the PET coordinate system through the registration matrix, including: The point coordinates in the CT coordinate system before radiotherapy are formed into a vector (X, Y, Z), which is multiplied by the registration matrix to realize the rotation and translation of the coordinate system and converted into the point coordinates in the PET coordinate system.

6. The lung tumor target area PET / CT and radiotherapy dose image extraction and processing method according to claim 1, characterized in that: If there is no valid data for the eight vertices of the smallest enclosing cuboid of the target object on the grid, all grid points on the six faces of the next smallest enclosing cuboid are selected as the source of fitting data. If the next smallest enclosing cuboid still cannot achieve the fitting task, a larger enclosing cuboid is selected, and so on, until the fitting task is achieved.

Images