Radiotherapy information determination device, storage medium and computer equipment

Abstract

The invention discloses a radiotherapy information determination device, a storage medium and computer equipment, relates to the technical field of image processing, and mainly aims to improve the image processing accuracy and the radiotherapy scheme determination accuracy. Comprising an information acquisition unit, a CT image processing unit and a radiotherapy information formulation unit. The information acquisition unit is used for responding to the tumor radiotherapy area division signal of the liver cancer patient and acquiring a multi-layer spiral CT image of the liver cancer patient about an affected tumor, radiotherapy demand information of the radiotherapy terminal for the liver cancer patient, operation demand information of the operation terminal for the liver cancer patient and patient feature information of the liver cancer patient; the CT image processing unit is used for segmenting a tumor focus area and a normal organ area around the tumor focus area from the multi-layer spiral CT image; and the radiotherapy information making unit is used for determining radiotherapy plan information of the liver cancer patient based on the radiotherapy demand information, the operation demand information, the tumor focus area, the normal organ area and the patient feature information.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a device for determining radiotherapy information, a storage medium and a computer device. Background Art

[0002] Radiotherapy is a common method of tumor treatment and is widely used in neoadjuvant treatment of head and neck tumors, liver cancer, pancreatic cancer, and rectal cancer. Planned radiotherapy of the primary tumor and nearby infiltrating lesions or regional metastatic lymph nodes can eliminate subclinical lesions and shrink the primary lesions, making it possible for patients who were originally unsuitable for surgery or who could not undergo surgery to undergo surgery, reducing the scope of surgery and better preserving the patient's postoperative physiological and living abilities. Radiotherapy can also reduce the vitality of tumor cells, reduce the chances of lymphatic and hematogenous metastasis of cancer cells, and reduce local implantation of cancer cells, thereby improving the cure rate. Based on this, it is crucial for doctors to reasonably formulate radiotherapy plan information.

[0003] At present, tumor images are usually processed based on the personal experience of the radiotherapist to determine the radiotherapy plan information. However, manually processing tumor images based on the personal experience of the radiotherapist will result in low accuracy of image processing. At the same time, formulating radiotherapy plan information based on the single factor of the radiotherapist's personal ideas will result in low accuracy of radiotherapy plan information. Summary of the invention

[0004] The present invention provides a device for determining radiotherapy information, a storage medium and a computer device, which are mainly capable of improving the accuracy of image processing and the accuracy of determining radiotherapy plans.

[0005] According to a first aspect of the present invention, there is provided a radiotherapy information determination device, comprising: an information acquisition unit, a CT image processing unit, and a radiotherapy information formulation unit;

[0006] The information acquisition unit is used to respond to the tumor radiotherapy area division signal of the liver cancer patient, obtain the multi-slice spiral CT image of the liver cancer patient's tumor, the radiotherapy requirement information of the liver cancer patient from the radiotherapy terminal, the surgery requirement information of the liver cancer patient from the surgery terminal, and the patient characteristic information of the liver cancer patient;

[0007] The CT image processing unit is used to segment the tumor lesion area and the normal organ area around the tumor lesion area from the multi-slice spiral CT image;

[0008] The radiotherapy information formulation unit is used to determine the radiotherapy plan information of the liver cancer patient based on the radiotherapy requirement information, the surgery requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information.

[0009] Optionally, the CT image processing unit includes a processing module and a first determining module; the CT image processing unit segments the tumor lesion area and the normal organ area around the tumor lesion area from the multi-slice spiral CT image, including:

[0010] The processing module is used to perform grayscale processing on the multi-slice spiral CT image to obtain a grayscale image corresponding to the multi-slice spiral CT image;

[0011] The first determination module is used to determine a growth starting pixel point in the tumor lesion area in the grayscale image, determine a preset neighborhood of the growth starting pixel point, and select neighborhood pixels whose pixel value difference with the growth starting pixel point satisfies a preset difference condition within the preset neighborhood, and use the area formed by the neighborhood pixels and the growth starting pixel point as the growth area;

[0012] The first determination module is specifically used to use each pixel point in the growth area as a new growth starting pixel point, continue to determine a new preset neighborhood corresponding to the new growth starting pixel point, and select new neighborhood pixels in the new preset neighborhood whose pixel value difference with the new growth starting pixel point meets a preset difference condition, and use the area formed by the new neighborhood pixels and the new growth starting pixel point as the new growth area, until no new neighborhood pixels are found in the new preset neighborhood, and determine the finally determined new growth area as the tumor lesion area, and determine the area around the tumor lesion area as the normal organ area.

[0013] Optionally, it further includes a division unit, an addition unit, an update unit, and a splicing unit; before the CT image processing unit segments the tumor lesion area and the normal organ area around the tumor lesion area from the multi-slice spiral CT image, it further includes:

[0014] The CT image processing unit may also be used to segment the grayscale image corresponding to the multi-slice spiral CT image into a plurality of non-overlapping sub-image blocks;

[0015] The division unit is used to determine the number of pixels of each gray level in each sub-image block, and divide the number of pixels of each gray level by the total number of pixels of the corresponding sub-image block to obtain the probability density of each gray level in each sub-image block;

[0016] The adding unit is used to sort the gray levels in ascending order, and take any gray level except the first gray level in the sorted gray levels as a current gray level, and add the probability density of each current gray level with the probability density of each gray level sorted before the current gray level to obtain the cumulative probability density corresponding to each current gray level;

[0017] The updating unit is used to multiply the cumulative probability density by the maximum gray level in the corresponding sub-image block to obtain a new gray level corresponding to each current gray level, and update each pixel value at each current gray level to a pixel value corresponding to the new gray level to obtain each sub-image block after equalization;

[0018] The splicing unit is used to determine the original position of each of the sub-image blocks in the grayscale image, and splice the equalized sub-image blocks according to the original position to obtain the pre-processed multi-slice spiral CT image.

[0019] Optionally, the radiotherapy information formulation unit includes: a trimming module, a second determination module, an acquisition module, and an expansion module; the radiotherapy information formulation unit determines the radiotherapy plan information of the liver cancer patient based on the radiotherapy requirement information, the surgical requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information, including:

[0020] The trimming module is used to trim the tumor lesion area and the normal organ area in the multi-slice spiral CT image based on the radiotherapy requirement information to obtain the trimmed tumor lesion area and the trimmed normal organ area;

[0021] The second determination module is used to determine the portal vein cancer thrombus area in the multi-slice spiral CT image based on the trimmed tumor lesion area and the trimmed normal organ area;

[0022] The second determination module is further used to determine the lesion area to be removed and the prohibited radiotherapy irradiation area in the multi-slice spiral CT image based on the surgical requirement information, the trimmed tumor lesion area, the trimmed normal organ area, and the portal vein cancer thrombus area;

[0023] The second determination module is further used to determine the initial radiotherapy area of ​​the liver cancer patient in the multi-slice spiral CT image based on the trimmed tumor lesion area, the trimmed normal organ area, the venous cancer thrombus area, the lesion area to be resected, and the radiotherapy irradiation prohibited area;

[0024] The acquisition module is used to acquire patient positioning information when acquiring CT images of the liver cancer patient;

[0025] The expansion module is used to determine the expansion information for the initial radiotherapy area based on the patient positioning information, and expand the initial radiotherapy area based on the expansion information to obtain the radiotherapy area for the liver cancer patient;

[0026] The second determination module is further used to determine the radiotherapy dose for the radiotherapy area based on the area size of the radiotherapy area and the patient characteristic information.

[0027] Optionally, the method further includes: a three-dimensional reconstruction unit and a fusion unit; after the radiotherapy information formulation unit determines the radiotherapy plan information of the liver cancer patient based on the radiotherapy requirement information, the surgery requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information, the method further includes:

[0028] The information acquisition unit is further used to obtain target area and actual radiation dose distribution information of the liver cancer patient after actual radiotherapy in response to a radiotherapy end signal of the liver cancer patient based on the radiotherapy plan information;

[0029] The three-dimensional reconstruction unit is used to perform three-dimensional reconstruction on the multi-slice spiral CT image to obtain a three-dimensional image corresponding to the multi-slice spiral CT image;

[0030] The fusion unit is used to align and fuse the target area after the actual radiotherapy, the actual radiation dose distribution information and the three-dimensional image to obtain the three-dimensional image with annotation information, wherein the annotation information includes a annotation frame of the target area after the actual radiotherapy and the actual radiation dose distribution information;

[0031] The radiotherapy information formulation unit is used to determine the surgical treatment planning information of the liver cancer patient based on the three-dimensional image with the annotation information.

[0032] Optionally, the multi-slice spiral CT image includes: a late arterial phase CT image and a venous phase CT image of the liver cancer patient;

[0033] The CT image processing unit segments a tumor lesion region and a normal organ region surrounding the tumor lesion region from the multi-slice spiral CT image, comprising:

[0034] The CT image processing unit is used to input the late arterial CT image into a first image segmentation model constructed based on a deep learning architecture to perform image segmentation, so as to obtain the hepatic artery region of the liver cancer patient;

[0035] The CT image processing unit is used to input the venous phase CT image into a second image segmentation model constructed based on a deep learning architecture to perform image segmentation to obtain the tumor lesion area, portal vein area, hepatic vein area, and bile duct area of ​​the liver cancer patient.

[0036] Optionally, the radiotherapy information formulation unit further includes: a cross module and a prediction module; the radiotherapy information formulation unit determines the radiotherapy plan information of the liver cancer patient based on the radiotherapy requirement information, the surgical requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information, including:

[0037] The second determination module is further used to respectively determine the radiotherapy feature vector corresponding to the radiotherapy requirement information, the surgery feature vector corresponding to the surgery requirement information, the tumor feature vector corresponding to the tumor lesion area, the organ feature vector corresponding to the normal organ area, and the patient feature vector corresponding to the patient feature information;

[0038] The cross module is used to perform cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a radiotherapy information cross vector;

[0039] The prediction module is used to input the radiotherapy information cross vector into a preset radiotherapy plan prediction model to perform radiotherapy plan prediction, so as to obtain the radiotherapy plan information of the liver cancer patient.

[0040] Optionally, the cross module includes: a cross submodule and a transformation submodule; the cross module performs cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a radiotherapy information cross vector, including:

[0041] The cross submodule is used to perform feature-level cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a feature cross vector;

[0042] The crossover submodule is further used to perform element-level crossover processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain an element crossover vector;

[0043] The cross submodule is further used to perform low-order cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a low-order cross vector;

[0044] The transformation submodule is used to transform the characteristic cross vector, the element cross vector, and the low-order cross vector to obtain a radiotherapy information cross vector.

[0045] According to a second aspect of the present invention, there is provided a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the following steps:

[0046] In response to a tumor radiotherapy area division signal of a liver cancer patient, obtaining a multi-slice spiral CT image of the liver cancer patient's tumor, radiotherapy requirement information of the liver cancer patient at a radiotherapy terminal, surgical requirement information of the liver cancer patient at a surgical terminal, and patient characteristic information of the liver cancer patient;

[0047] Segmenting a tumor lesion region and a normal organ region surrounding the tumor lesion region from the multi-slice spiral CT image;

[0048] Based on the radiotherapy requirement information, the surgical requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information, the radiotherapy plan information of the liver cancer patient is determined.

[0049] According to a fourth aspect of the present invention, there is provided a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the following steps are implemented:

[0050] In response to a tumor radiotherapy area division signal of a liver cancer patient, obtaining a multi-slice spiral CT image of the liver cancer patient's tumor, radiotherapy requirement information of the liver cancer patient at a radiotherapy terminal, surgical requirement information of the liver cancer patient at a surgical terminal, and patient characteristic information of the liver cancer patient;

[0051] Segmenting a tumor lesion region and a normal organ region surrounding the tumor lesion region from the multi-slice spiral CT image;

[0052] Based on the radiotherapy requirement information, the surgical requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information, the radiotherapy plan information of the liver cancer patient is determined.

[0053] According to a radiotherapy information determination device, storage medium and computer device provided by the present invention, compared with the current method of processing tumor images to determine radiotherapy plans based only on the personal experience of radiotherapy doctors, the present invention constructs a radiotherapy information determination device, obtains multi-slice spiral CT images of liver cancer patients, radiotherapy requirement information for liver cancer patients at a radiotherapy terminal, surgical requirement information for liver cancer patients at a surgical terminal, and patient characteristic information of liver cancer patients through an information acquisition unit in the determination device, and segments tumor lesion areas and normal organ areas from multi-slice spiral CT images through a CT image processing unit in the determination device. Finally, the radiotherapy information formulation unit in the determination device performs a comprehensive analysis of the radiotherapy requirement information, surgical requirement information, tumor lesion areas, normal organ areas, and patient characteristic information, and determines radiotherapy plan information based on the analysis results. Therefore, the present invention segments the tumor lesion areas and normal organ areas in the multi-slice spiral CT images through the CT image processing unit, which can avoid wasted time and segmentation errors caused by manual segmentation of the tumor lesion areas and normal organ areas, thereby improving the processing efficiency and accuracy of CT images. At the same time, when formulating radiotherapy plan information, the embodiment of the present invention utilizes a radiotherapy information determination device to perform a comprehensive analysis of the radiotherapy physician's radiotherapy demand information, the surgeon's surgical demand information, the tumor lesion area, the normal organ area, and the patient's characteristic information. This enables the formulated radiotherapy plan information to meet both the radiotherapy needs of the radiotherapy physician and the surgical needs of the surgeon during the operation, while also making the determined radiotherapy information suitable for the personalized physical sign information of the cancer patient, thereby making the formulated radiotherapy plan information more accurate and providing a better reference basis for doctors. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0055] Figure 1 A schematic diagram showing a process of determining radiotherapy information by a device for determining radiotherapy information provided by an embodiment of the present invention;

[0056] Figure 2 A schematic diagram showing a flow chart of another apparatus for determining radiotherapy information provided by an embodiment of the present invention for determining radiotherapy information;

[0057] Figure 3 A schematic diagram showing the structure of a device for determining radiotherapy information provided by an embodiment of the present invention;

[0058] Figure 4 A schematic diagram showing the structure of another apparatus for determining radiotherapy information provided by an embodiment of the present invention;

[0059] Figure 5 A schematic diagram of the physical structure of a computer device provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0060] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0061] At present, the method of processing tumor images to determine radiotherapy plans based solely on the radiotherapist's personal experience will result in low accuracy in image processing. At the same time, formulating radiotherapy plan information based solely on the single factor of the radiotherapist's personal ideas will result in low accuracy in determining radiotherapy plan information.

[0062] In order to solve the above problems, an embodiment of the present invention provides a process of determining radiotherapy information by a device for determining radiotherapy information, such as Figure 1 As shown, the device includes: an information acquisition unit, a CT image processing unit, and a radiotherapy information formulation unit; wherein the information acquisition unit, the CT image processing unit, and the radiotherapy information formulation unit are combined with each other to implement the following radiotherapy information determination process:

[0063] 101. In response to the tumor radiotherapy area division signal of the liver cancer patient, the information acquisition unit acquires the multi-slice spiral CT image of the liver cancer patient's tumor, the radiotherapy requirement information of the liver cancer patient at the radiotherapy terminal, the surgical requirement information of the liver cancer patient at the surgical terminal, and the patient characteristic information of the liver cancer patient, and sends the multi-slice spiral CT image to the CT image processing unit, and sends the radiotherapy requirement information, the surgical requirement information, and the patient characteristic information to the radiotherapy information formulation unit.

[0064] Among them, radiotherapy demand information refers to the radiotherapy implementation method, pre-implemented radiotherapy technology, pre-planned radiotherapy range and other information planned by the radiotherapy doctor for liver cancer patients; surgical demand information refers to the surgical implementation method, pre-implemented surgical technology, pre-planned surgical resection range, prohibited radiotherapy area and other information planned by the surgeon for liver cancer patients; patient characteristic information includes: age, physical condition, previous medical history and other information of liver cancer patients; multi-slice spiral CT images are obtained by pre-injecting contrast agent into liver cancer patients and then performing spiral scans on liver cancer patients. It should be noted that the scanning range should cover the liver and surgical site of liver cancer patients, and can be extended by at least 5 cm in the direction of the patient's head and feet if necessary.

[0065] For the embodiment of the present invention, when it is necessary to formulate radiotherapy plan information, it is first necessary to determine the radiotherapy area, and at this time, the tumor radiotherapy area determination signal is triggered. When the information acquisition unit receives the tumor radiotherapy area division signal, the multi-slice spiral CT image of the liver cancer patient, the radiotherapy requirement information for the liver cancer patient at the radiotherapy terminal, the surgical requirement information for the liver cancer patient at the surgical terminal, and the patient characteristic information of the liver cancer patient are obtained. It should be noted that the radiotherapy requirement information, surgical requirement information, and patient characteristic information obtained in this embodiment are not private data, but data that can be obtained in the process of determining the radiotherapy information (non-private data). Furthermore, after the information acquisition unit obtains the above information, the multi-slice spiral CT image is sent to the CT image processing unit, and the remaining information, including the radiotherapy requirement information, the surgical requirement information, and the patient characteristic information, is sent to the radiotherapy information formulation unit.

[0066] 102. The CT image processing unit receives the multi-slice spiral CT image sent by the information acquisition unit, segments the tumor lesion area and the normal organ area surrounding the tumor lesion area from the multi-slice spiral CT image, and sends the tumor lesion area and the normal organ area to the radiotherapy information formulation unit.

[0067] Among them, multi-slice spiral CT images include late arterial phase CT images and venous phase CT images of patients with liver cancer; the tumor lesion area refers to the area in the liver where the tumor grows, including but not limited to the left lobe, right lobe, left lateral lobe, right posterior lobe, etc.; the normal organ area refers to the intrahepatic bile duct, intrahepatic blood vessels, portal vein, hepatic vein, bile duct, biliary system, stomach, duodenum, pancreas, etc.

[0068] For the embodiment of the present invention, after the CT image processing unit receives the multi-slice spiral CT image sent by the information acquisition unit, it is necessary to segment the multi-slice spiral CT image. Based on this, step 102 specifically includes: the CT image processing unit is used to input the late arterial phase CT image into a first image segmentation model constructed based on a deep learning architecture for image segmentation to obtain the hepatic artery region of the liver cancer patient; the CT image processing unit is used to input the venous phase CT image into a second image segmentation model constructed based on a deep learning architecture for image segmentation to obtain the tumor lesion region, portal vein region, hepatic vein region, hepatic artery region, and bile duct region of the liver cancer patient.

[0069] Specifically, the CT image processing unit inputs the late arterial phase CT image into a first image segmentation model constructed based on a deep learning architecture for image segmentation, and obtains normal organ regions such as the hepatic artery region of liver cancer patients. At the same time, the CT image processing unit inputs the venous phase CT image into a second image segmentation model constructed based on a deep learning architecture for image segmentation, and obtains the tumor lesion region of liver cancer patients, as well as normal organ tissue regions such as the portal vein region, hepatic vein region, and bile duct region, and performs liver segmentation. In order to improve the segmentation accuracy of the image segmentation model in the embodiment of the present invention, before using the image segmentation model to segment the CT image, it is necessary to construct the image segmentation model. Taking the construction process of the first image segmentation model as an example, the specific construction method includes: constructing an initial image segmentation model, wherein the model architecture of the initial image segmentation model can be a deep learning architecture; obtaining a sample data set, wherein the sample data set includes a sample arterial late CT image with annotation information of a sample liver cancer object, wherein the annotation information is an annotation box of a tumor lesion area and an annotation box of a normal organ area; dividing the sample data set into a training set and a test set, using the training set to train the initial image segmentation model, and using the test set to test the trained initial image segmentation model, and determining the initial image segmentation model whose test results meet the test conditions as the first image segmentation model. Similarly, according to the above method, the second image segmentation model can be trained and constructed. Therefore, the embodiment of the present invention uses a CT image processing unit to segment the tumor lesion area and the normal organ area in the multi-slice spiral CT image, which can avoid the time wasted and segmentation errors caused by manual segmentation of the tumor lesion area and the normal organ area, so that the present invention can improve the processing efficiency and processing accuracy of CT images.

[0070] 103. The radiotherapy information formulation unit receives the radiotherapy requirement information, surgical requirement information, and patient characteristic information sent by the information acquisition unit, and receives the tumor lesion area and normal organ area sent by the CT image processing unit, and determines the radiotherapy plan information of the liver cancer patient based on the radiotherapy requirement information, surgical requirement information, tumor lesion area, normal organ area, and patient characteristic information.

[0071] Among them, the radiotherapy plan information includes the radiotherapy area and the radiotherapy dose for the radiotherapy area; the radiation dose refers to the amount of radiation energy given to the patient's body during the radiotherapy process; the radiotherapy information formulation unit includes a trimming module, a second determination module, an acquisition module, and an expansion module.

[0072] For the embodiment of the present invention, after receiving the tumor lesion area and the normal organ area sent by the CT image processing unit, and receiving the radiotherapy requirement information, the surgical requirement information, and the patient characteristic information sent by the information acquisition unit, the radiotherapy information formulation unit needs to conduct a comprehensive analysis of the above information and determine the radiotherapy plan information of the liver cancer patient according to the analysis results. Based on this, the method for the radiotherapy information formulation unit to formulate the radiotherapy plan information specifically includes: the trimming module is used to trim the tumor lesion area and the normal organ area in the multi-slice spiral CT image based on the radiotherapy requirement information to obtain the trimmed tumor lesion area and the trimmed normal organ area; the second determination module is used to determine the portal vein cancer thrombus area in the multi-slice spiral CT image based on the trimmed tumor lesion area and the trimmed normal organ area; the second determination module is also used to trim the tumor lesion area and the trimmed normal organ area based on the surgical requirement information The lesion area, the trimmed normal organ area, and the portal vein cancer thrombus area are used to determine the lesion area to be removed and the prohibited radiotherapy irradiation area in the multi-slice spiral CT image; the second determination module is also used to determine the initial radiotherapy area of ​​the liver cancer patient in the multi-slice spiral CT image based on the trimmed tumor lesion area, the trimmed normal organ area, the venous cancer thrombus area, the lesion area to be removed, and the prohibited radiotherapy irradiation area; the acquisition module is used to acquire the patient positioning information when the CT image of the liver cancer patient is acquired; the expansion module is used to determine the expansion information for the initial radiotherapy area based on the patient positioning information, and expand the initial radiotherapy area based on the expansion information to obtain the radiotherapy area of ​​the liver cancer patient; the second determination module is also used to determine the radiotherapy dose for the radiotherapy area based on the area size of the radiotherapy area and the patient characteristic information.

[0073] Specifically, the trimming module in the radiotherapy information formulation unit reviews or modifies the segmentation results of the tumor and normal tissues and organs in the determination device according to the radiotherapy demand information, and trims the tumor lesion area and the normal organ area according to the recognition result to obtain the corrected tumor lesion area and the normal organ area, and sends the corrected tumor lesion area and the normal organ area to the second determination unit. After receiving the corrected tumor lesion area and the normal organ area, the second determination unit identifies and outlines the portal vein cancer thrombus area in the corrected tumor lesion area and the normal organ area, and then plans the surgical path according to the surgical demand information, the trimmed tumor lesion area, the trimmed normal organ area, and the portal vein cancer thrombus area, and marks the lesions to be removed, the hepatic artery segments and normal tissues that should be avoided from irradiation, and the normal tissues or diseased tissues that do not need to be irradiated in the CT image according to the tumor condition and the residual liver volume that needs to be preserved, that is, the identification The lesion area to be removed and the area prohibited from radiotherapy are determined, and the initial radiotherapy area of ​​the liver cancer patient is determined in the multi-slice spiral CT image according to the trimmed tumor lesion area, the trimmed normal organ area, the venous cancer thrombus area, the lesion area to be removed and the area prohibited from radiotherapy, and the initial radiotherapy area is sent to the expansion module. At the same time, the acquisition module is used to determine the patient positioning information when the CT image of the liver cancer patient is acquired, that is, the position between the patient and the treatment bed, including the body position and posture, and the positioning information is sent to the expansion module. After receiving the patient positioning information, the expansion module determines the expansion information of the initial radiotherapy area according to the patient positioning information, and expands the initial radiotherapy area according to the expansion information to obtain the radiotherapy area, and sends the radiotherapy area to the second determination module. After receiving the radiotherapy area, the second determination module determines the radiotherapy dose for the radiotherapy area according to the area size of the radiotherapy area and the patient's age, previous medical history and other characteristics. Therefore, the embodiment of the present invention can ensure that the radiotherapy dose can fully cover the tumor and its potential growth area by expanding the initial radiotherapy area, thereby improving the tumor control rate and avoiding the error in determining the radiotherapy area caused by the patient's positioning deviation.

[0074] For the embodiment of the present invention, after the radiotherapy information formulation unit receives the radiotherapy requirement information, surgical requirement information, and patient characteristic information sent by the information acquisition unit, and receives the tumor lesion area and normal organ area sent by the CT image processing unit, it combines the radiotherapy requirement information, surgical requirement information, and patient characteristic information to reasonably plan the tumor lesion area and normal organ area to obtain the radiotherapy area and the radiotherapy dose for the radiotherapy area. When formulating radiotherapy plan information, the embodiment of the present invention uses the radiotherapy information determination device to comprehensively analyze the radiotherapy requirement information of the radiotherapy doctor, the surgical requirement information of the surgeon, the tumor lesion area, the normal organ area, and the patient characteristic information, so that the formulated radiotherapy plan information can meet both the radiotherapy needs of the radiotherapy doctor and the surgical needs of the surgeon during the operation, and at the same time make the determined radiotherapy information suitable for the personalized physical sign information of the cancer patient, so that the formulated radiotherapy plan information can meet both the needs of the radiotherapy doctor and the needs of the surgeon, so as to facilitate application in various subsequent scenarios.

[0075] According to a device for determining radiotherapy information provided by the present invention, compared with the current method of processing tumor images to determine radiotherapy plans based solely on the personal experience of radiotherapists, the present invention constructs a device for determining radiotherapy information, obtains multi-slice spiral CT images of liver cancer patients, radiotherapy requirement information for liver cancer patients at a radiotherapy terminal, surgical requirement information for liver cancer patients at a surgical terminal, and patient characteristic information of liver cancer patients through an information acquisition unit in the determination device, and segments tumor lesion areas and normal organ areas from multi-slice spiral CT images through a CT image processing unit in the determination device. Finally, a comprehensive analysis of the radiotherapy requirement information, surgical requirement information, tumor lesion areas, normal organ areas, and patient characteristic information is performed through a radiotherapy information formulation unit in the determination device, and radiotherapy plan information is determined based on the analysis results. Therefore, the present invention segments the tumor lesion areas and normal organ areas in the multi-slice spiral CT images through the CT image processing unit, thereby avoiding the time wasted and segmentation errors caused by manual segmentation of the tumor lesion areas and normal organ areas, thereby improving the processing efficiency and accuracy of CT images. At the same time, when formulating radiotherapy plan information, the embodiment of the present invention utilizes a radiotherapy information determination device to perform a comprehensive analysis of the radiotherapy physician's radiotherapy demand information, the surgeon's surgical demand information, the tumor lesion area, the normal organ area, and the patient's characteristic information. The formulated radiotherapy plan information can meet both the radiotherapy needs of the radiotherapy physician and the surgical needs of the surgeon during the operation, and at the same time make the determined radiotherapy information suitable for the personalized physical sign information of the cancer patient, thereby enabling the formulated radiotherapy plan information to meet the needs of multiple parties.

[0076] Further, in order to better illustrate the process of determining radiotherapy information using the above-mentioned device for determining radiotherapy information, as a refinement and extension of the above-mentioned embodiment, the embodiment of the present invention provides another process of determining radiotherapy information using the device for determining radiotherapy information, such as Figure 2 As shown, including:

[0077] 201. An information acquisition unit, configured to respond to a tumor radiotherapy area division signal of a liver cancer patient to obtain a multi-slice spiral CT image of the liver cancer patient's tumor, radiotherapy requirement information for the liver cancer patient from a radiotherapy terminal, surgery requirement information for the liver cancer patient from a surgery terminal, and patient characteristic information of the liver cancer patient.

[0078] For this embodiment of the present invention, when the information acquisition unit receives a tumor radiotherapy area division signal, it will acquire a multi-slice spiral CT image, radiotherapy requirement information for liver cancer patients from a radiotherapy terminal, surgical requirement information for liver cancer patients from a surgical terminal, and patient characteristic information of liver cancer patients, and send the CT image to a CT image processing unit, and send the radiotherapy requirement information, surgical requirement information, and patient characteristic information to a radiotherapy information formulation unit.

[0079] Furthermore, before or after receiving the multi-slice spiral CT image, the CT image processing unit needs to preprocess the CT image first before segmenting the CT image in order to improve the segmentation accuracy of the multi-slice spiral CT image. Based on this, the radiotherapy information determination device also includes a division unit, an addition unit, an update unit, and a splicing unit. The specific use of the above units to implement the preprocessing process of the CT image includes: the CT image processing unit can also be used to divide the grayscale image corresponding to the multi-slice spiral CT image into multiple non-overlapping sub-image blocks; the division unit is used to respectively determine the number of pixels of each gray level in each of the sub-image blocks, and divide the number of pixels of each gray level by the total number of pixels of the corresponding sub-image block to obtain the probability density of each gray level in each of the sub-image blocks; the addition unit is used to calculate the probability density of each gray level according to the gray level. The gray levels are sorted in ascending order, and any gray level except the first gray level in the sorted gray levels is taken as a current gray level, and the probability density of each current gray level is added with the probability density of each gray level sorted before the current gray level to obtain the cumulative probability density corresponding to each current gray level; the updating unit is used to multiply the cumulative probability density with the maximum gray level in the corresponding sub-image block to obtain a new gray level corresponding to each current gray level, and update each pixel value under each current gray level to the pixel value corresponding to the new gray level to obtain each equalized sub-image block; the splicing unit is used to determine the original position of each sub-image block in the gray image, and splice the equalized sub-image blocks according to the original position to obtain the preprocessed multi-slice spiral CT image.

[0080] Specifically, after receiving the multi-slice spiral CT image, the CT image processing unit first performs grayscale processing on the multi-slice spiral CT image to obtain a grayscale image, and divides the grayscale image into a plurality of non-overlapping sub-image blocks, wherein the number of sub-image blocks is not specifically limited in the embodiment of the present invention, and then sends the grayscale image to the division unit. After receiving each sub-image block, the division unit performs equalization processing on each sub-image block. The specific equalization processing process is as follows: taking sub-image block a as an example, first use the division unit to determine each grayscale level corresponding to sub-image block a, such as 0 to 255, 0-2 55 are the gray values ​​in sub-image block a respectively, and the number of pixels at each gray level is determined respectively, such as the number of pixels with a gray value of 0, the number of pixels with a gray value of 255, etc. If the number of pixels with a gray level of 255 is 2000, and the total number of pixels corresponding to sub-image block a is 5000, then the probability density of gray level 255 in sub-image block a is m=2000÷5000=0.4. According to the above method, the probability density corresponding to each gray level in sub-image block a can be determined, such as determining that the probability density corresponding to gray level A is 0.1, and the probability density corresponding to gray level B is 0.2. , the probability density corresponding to gray level C is 0.3, and the probability density corresponding to gray level D is 0.4. Then, the probability density corresponding to each gray level is sent to the adding unit. After receiving the probability density corresponding to each gray level under each sub-image block, the adding unit first sorts each gray level, such as gray level A, gray level B, gray level C, and gray level D in descending order. Then, the probability density corresponding to gray level B is added to the probability density corresponding to gray level A by the adding unit, that is, 0.3 is determined as the cumulative probability density corresponding to gray level B. Similarly, the probability density corresponding to gray level C is The probability density is added to the probability density corresponding to gray level A and gray level B, that is, 0.6 is determined as the cumulative probability density corresponding to gray level C, and the probability density corresponding to gray level D is added to the probability density corresponding to gray level A, gray level B, and gray level C, that is, 1 is determined as the cumulative probability density corresponding to gray level D. If the maximum gray level corresponding to sub-image block a is 255, the maximum gray level 255 is multiplied by the cumulative probability density 0.3 corresponding to gray level B, and the obtained 76.5 is the new gray level corresponding to gray level B. The maximum gray level 255 is multiplied by the cumulative probability density 0.6, and the obtained 153 is the new gray level corresponding to the gray level C. Therefore, according to the above method, the new gray level corresponding to each gray level in each image block can be obtained, and each new gray level is sent to the updating unit. After receiving the new gray level, the updating unit uses the pixel value corresponding to the new gray level to update the pixel value corresponding to each pixel point under the corresponding gray level to obtain each equalized sub-image block, and sends each equalized sub-image block to the splicing unit. After receiving each equalized sub-image block, the splicing unit determines the position of the original sub-image block corresponding to the equalized sub-image block, and splices the equalized sub-image blocks according to the position to obtain the preprocessed multi-slice spiral CT image. Therefore, the embodiment of the present invention pre-processes the multi-slice spiral CT image by equalization. Since the equalization process redistributes the pixel values ​​of the image, the brightness distribution of the image is more uniform, so that the originally too dark or too bright areas will be properly adjusted, and the contrast of the image is significantly enhanced, which helps to better display the tissue structure and the lesion area. At the same time, through the equalization process, the details and edges of the image are enhanced, making the tiny structures or lesions in the image more clearly visible. Finally, the image after equalization is clearer and brighter in visual effect, and the overall quality is improved, thereby improving the segmentation effect of the CT image.

[0081] Furthermore, after the CT image is preprocessed to obtain the preprocessed multi-slice spiral CT image, the splicing unit sends the preprocessed multi-slice spiral CT image to the CT image processing unit, and the multi-slice spiral CT image is segmented by the CT image processing unit. At the same time, the information acquisition unit sends the radiotherapy requirement information, the surgical requirement information, and the patient characteristic information to the radiotherapy information formulation unit.

[0082] 202. A processing module, configured to perform grayscale processing on the multi-slice spiral CT image to obtain a grayscale image corresponding to the multi-slice spiral CT image.

[0083] Specifically, the CT image processing unit includes a processing module and a first determination module. The processing module is used to receive the multi-slice spiral CT image sent by the information acquisition unit, and perform grayscale processing on the multi-slice spiral CT image to obtain a grayscale image corresponding to the multi-slice spiral CT image, and send the grayscale image to the first determination module.

[0084] 203. A first determination module is used to determine a growth starting pixel point in the tumor lesion area in the grayscale image, determine a preset neighborhood of the growth starting pixel point, and select neighborhood pixels whose pixel value difference with the growth starting pixel point satisfies a preset difference condition within the preset neighborhood, and use the area formed by the neighborhood pixels and the growth starting pixel point as the growth area.

[0085] 204. A first determination module is specifically used to use each pixel point in the growth area as a new growth starting pixel point, continue to determine a new preset neighborhood corresponding to the new growth starting pixel point, and select a new neighborhood pixel point in the new preset neighborhood whose pixel value difference with the new growth starting pixel point meets a preset difference condition, and use the area formed by the new neighborhood pixel point and the new growth starting pixel point as the new growth area, until no new neighborhood pixel point is found in the new preset neighborhood, and determine the finally determined new growth area as the tumor lesion area, and determine the area around the tumor lesion area as the normal organ area.

[0086] The preset neighborhood is set according to actual needs; the preset difference condition is set according to actual needs. Specifically, after receiving the grayscale image, the first determination module can firstly use an adaptive algorithm to automatically determine the growth starting pixel point. For example, by traversing the entire CT image, the local grayscale value or gradient value of each pixel is calculated, and then the pixel points that meet specific conditions are selected as the growth starting pixel points, and then the pixel points similar to the growth starting pixel points are selected from the various pixels in the preset neighborhood, that is, the neighborhood pixel points. Specifically, the similarity criterion can be formulated according to the grayscale value, gradient value or other characteristics of the pixel points. For example, the pixels whose grayscale value differences are within a certain range can be set as similar pixels. Then, the area formed by the growth starting pixel points and similar pixels (neighborhood pixel points) is used as the growth area, and each pixel point in the growth area is used as a new growth starting pixel point, and a new preset neighborhood corresponding to the growth starting pixel point is determined. According to the above method, new neighborhood pixel points similar to the new growth starting pixel point are selected in the new preset neighborhood, and the area formed by the neighborhood pixel points and the new growth starting pixel point is determined as the new growth area, thereby continuously repeating the above process, that is, starting from the new growth starting point, the surrounding pixels are gradually merged according to the similarity criterion. After each merge, the characteristic values ​​of the growth area (such as grayscale mean, gradient mean, etc.) need to be recalculated in order to compare with the new pixel points. When no more pixels meet the similarity criterion, the regional growth process terminates. At this time, the final new growth area obtained is the segmented tumor area, and the area around the tumor area is the normal organ area. The embodiment of the present invention segments the CT image by determining the growth area. Since the growth area method can process medical images with complex grayscale distribution, such as soft tissue, bone and lesion areas in CT images, this method is not limited by the size and shape of the image and can be flexibly applied to various CT image segmentation tasks. The growth area method is based on the similarity of adjacent pixels for segmentation, and can accurately identify and segment the target area. In CT images, structures such as tumors and blood vessels usually have obvious grayscale differences from surrounding tissues. The growth area method can use these differences to achieve accurate segmentation, thereby helping to reduce subjectivity and errors in the segmentation process and improve segmentation efficiency and accuracy.

[0087] Furthermore, the first determination module in the CT image processing unit sends the tumor lesion area and the normal organ area to the radiotherapy information formulation unit.

[0088] 205. A radiotherapy information formulation unit is used to determine the radiotherapy plan information of the liver cancer patient based on the radiotherapy requirement information, the surgical requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information.

[0089] Among them, the radiotherapy information formulation unit also includes: a cross module and a prediction module.

[0090] For the embodiment of the present invention, after receiving the tumor lesion area and the normal organ area sent by the first determination module in the CT image processing unit, and receiving the radiotherapy requirement information, surgical requirement information, and patient characteristic information sent by the information acquisition unit, the radiotherapy information formulation unit needs to perform a comprehensive analysis of the above information and determine the radiotherapy plan information of the liver cancer patient according to the analysis results. Based on this, step 205 specifically includes: the second determination module is also used to respectively determine the radiotherapy feature vector corresponding to the radiotherapy requirement information, the surgical feature vector corresponding to the surgical requirement information, the tumor feature vector corresponding to the tumor lesion area, the organ feature vector corresponding to the normal organ area, and the patient feature vector corresponding to the patient characteristic information; the cross module is used to cross-process the radiotherapy feature vector, the surgical feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a radiotherapy information cross vector; the prediction module is used to input the radiotherapy information cross vector into a preset radiotherapy plan prediction model to perform radiotherapy plan prediction to obtain the radiotherapy plan information of the liver cancer patient. Among them, the cross module includes: a cross submodule and a transformation submodule; the method for the cross module to perform cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector includes: the cross submodule is used to perform feature-level cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a feature cross vector; the cross submodule is also used to perform element-level cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain an element cross vector; the cross submodule is also used to perform low-order cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a low-order cross vector; the transformation submodule is used to perform transformation processing on the feature cross vector, the element cross vector, and the low-order cross vector to obtain a radiotherapy information cross vector.

[0091] Specifically, in order to improve the prediction accuracy of the preset radiotherapy plan prediction model, the embodiment of the present invention first needs to train and construct the preset radiotherapy plan prediction model. Based on this, the radiotherapy information determination device also enters the construction unit, and the preset radiotherapy plan prediction model is constructed using the construction unit. The specific execution process of the construction unit is as follows: construct an initial prediction model and obtain a sample data set, wherein the sample data set contains sample radiotherapy demand information, sample surgical demand information, sample tumor lesion area, sample normal organ area, sample patient feature information, and actual radiotherapy plan information for sample liver cancer patients whose radiotherapy effects meet radiotherapy requirements and surgical requirements; divide the sample data set into a training set and a test set, use the training set to train the initial prediction model, and use the test set to test the trained initial prediction model, and finally determine the trained initial prediction model whose test results meet the test conditions as the preset radiotherapy plan prediction model. Thus, by pre-training and constructing the preset radiotherapy plan prediction model, the prediction accuracy of the preset radiotherapy plan prediction model for radiotherapy plan information can be improved.

[0092] Specifically, after constructing a preset radiotherapy plan prediction model with high prediction accuracy, it is necessary to use the model to predict the radiotherapy plan information of liver cancer patients. Based on this, after receiving the radiotherapy requirement information, surgical requirement information, tumor lesion area, normal organ area, and patient characteristic information, the radiotherapy information formulation unit first starts the second determination module in the radiotherapy information formulation unit to use word embedding and other methods to respectively determine the radiotherapy feature vector corresponding to the radiotherapy requirement information, the surgical feature vector corresponding to the surgical requirement information, the tumor feature vector corresponding to the tumor lesion area, the organ feature vector corresponding to the normal organ area, and the patient feature vector corresponding to the patient characteristic information. Since the radiotherapy requirement information, surgical requirement information, tumor lesion area, normal organ area, and patient characteristic information are different fields, the processing method is to embed the data of the five fields into vectors of the same latitude. For example, if the radiotherapy feature vector is (a1, a2), the surgery feature vector is (b1, b2), the tumor feature vector is (c1, c2), the organ feature vector is (d1, d2), and the patient feature vector is (e1, e2), the specific cross processing method includes: performing feature-level crossover between different feature vectors, that is, performing Hadamard product on all elements between vectors, and then performing convolution transformation under certain weights to obtain a feature crossover vector of f(w*(a1*b1*c1*d1*e1,a2*b2*c2*d2*e2,a3*b3*c3*d3*e3)); at the same time, performing element-level crossover on all feature vector data, that is, performing Hadamard product on each element between vectors Afterwards, assign different weight values ​​to the results of each multiplication, and then perform linear transformation, and the obtained element cross vector is f(w1*a1*b1*c1*d1*e1,w2*a2*b2*c2*d2*e2,w3*a3*b3*c3*d3*e3); in addition, all feature vectors are subjected to low-order cross processing, and weight coefficients are assigned to the results after cross processing, and then linear transformation is performed, and the obtained low-order cross vector is f(w2(a1,a2,b1,b2,c1,c2,d1,d2,e1,e2)); finally, the above feature cross vectors, element cross vectors, and low-order cross vectors are combined together, and the preset function is used for transformation processing to obtain the radiotherapy information cross vector. The preset function here can be set according to the actual situation, and this embodiment does not limit this. It should be noted that the above examples are only illustrative and do not limit the embodiments of the present application.Therefore, by cross-processing the radiotherapy feature vector, surgery feature vector, tumor feature vector, organ feature vector, and patient feature vector, different features can be automatically or explicitly combined to generate new feature combinations. These combined features may contain complex nonlinear relationships between the original features, so that the model can capture more detailed and rich information in the data, that is, it can make full use of the relationship between various data, extract more implicit features, and take into account both high-order and low-order processing, so that data utilization is more complete, and the radiotherapy plan information generation results obtained later are more accurate, meeting the needs of actual application scenarios.

[0093] Furthermore, after determining the radiotherapy information cross vector, the radiotherapy information cross vector is input into a preset radiotherapy plan prediction model for radiotherapy plan prediction, and the preset radiotherapy plan prediction model can output radiotherapy plan information for liver cancer patients. The embodiment of the present invention predicts radiotherapy plan information through a model, which can improve the determination efficiency and accuracy of the radiotherapy plan.

[0094] Furthermore, after the radiotherapy plan information is formulated, the radiotherapy terminal can realize multi-scenario application according to the radiotherapy plan information. After application, it is also necessary to formulate surgical planning information, that is, surgical treatment planning information. Specifically, the surgical planning information determination process can be realized through the three-dimensional reconstruction unit and the fusion unit in the radiotherapy information determination device, including: the information acquisition unit is also used to respond to the radiotherapy end signal of the liver cancer patient based on the radiotherapy plan information, and obtain the target area and actual radiation dose distribution information of the liver cancer patient after actual radiotherapy; the three-dimensional reconstruction unit is used to perform three-dimensional reconstruction on the multi-slice spiral CT image to obtain the three-dimensional image corresponding to the multi-slice spiral CT image; the fusion unit is used to align and fuse the target area after actual radiotherapy and the actual radiation dose distribution information with the three-dimensional image to obtain the three-dimensional image with annotation information, wherein the annotation information includes the annotation box of the target area after actual radiotherapy and the actual radiation dose distribution information; the radiotherapy information formulation unit is also used to determine the surgical treatment planning information of the liver cancer patient based on the three-dimensional image with annotation information.

[0095] Specifically, after the application of radiotherapy plan information, it is also necessary to specify surgical planning information for the scene after the application of radiotherapy information. Based on this, a three-dimensional reconstruction unit is also introduced in the radiotherapy information determination device. When the information acquisition unit receives the radiotherapy end information, the target area and actual radiation dose distribution information of the liver cancer patient after actual radiotherapy are obtained, and the three-dimensional reconstruction unit is used to perform three-dimensional reconstruction on the multi-slice spiral CT image of the liver cancer patient to obtain a three-dimensional image. Then, according to the characteristics of the target area and the three-dimensional image after actual radiotherapy, a suitable image registration algorithm is selected, such as feature-based registration, grayscale-based registration, and transformation-based registration. The registration is performed, that is, the target area image and the three-dimensional image are registered to make them consistent in three-dimensional space. Calculate indicators such as registration error and overlap to evaluate the quality of the registration result. If the quality does not meet the requirements, it is necessary to readjust the registration parameters or select other registration algorithms. At the same time, a fusion operation is performed according to the weighted average method, the maximum and minimum method, and the multi-resolution fusion method, that is, the selected fusion method is used to fuse the actual target area after radiotherapy, the actual radiation dose distribution information, and the three-dimensional image to generate a comprehensive image containing all information, that is, a three-dimensional image with annotated information. At the same time, according to the quality of the fusion result, the fusion parameters, such as weights, thresholds, etc., are adjusted to obtain the best fusion effect. Finally, according to the three-dimensional image with annotated information, the position, shape, and size of the target area after radiotherapy are accurately identified, and the anatomical relationship between the target area and the surrounding normal tissues is determined. The radiotherapy dose distribution image is analyzed to understand the radiotherapy dose received by the target area and the irradiation dose received by the surrounding normal tissues, and the killing effect of radiotherapy on the target area and the potential damage to the normal tissues are evaluated. Then, according to the position of the target area and the anatomical relationship of the surrounding tissues, a suitable surgical path is selected, and the scope of surgical resection is determined in combination with the changes in the target area after radiotherapy, so as to obtain surgical treatment planning information for liver cancer patients after radiotherapy. The embodiment of the present invention formulates a surgical planning plan by registering and fusing the three-dimensional images, which can help doctors understand the relationship between the lesion area and the surrounding structures more comprehensively before surgery, so as to select the best surgical path.

[0096] According to another radiotherapy information determination device provided by the present invention, compared with the current method of processing tumor images to determine radiotherapy plans based solely on the personal experience of radiotherapy doctors, the present invention constructs a radiotherapy information determination device, obtains multi-slice spiral CT images of liver cancer patients, radiotherapy requirement information for liver cancer patients at radiotherapy terminals, surgical requirement information for liver cancer patients at surgical terminals, and patient characteristic information of liver cancer patients through an information acquisition unit in the determination device, and segments tumor lesion areas and normal organ areas from multi-slice spiral CT images through a CT image processing unit in the determination device. Finally, the radiotherapy information formulation unit in the determination device performs a comprehensive analysis of the radiotherapy requirement information, surgical requirement information, tumor lesion areas, normal organ areas, and patient characteristic information, and determines radiotherapy plan information based on the analysis results. Therefore, the present invention segments the tumor lesion areas and normal organ areas in the multi-slice spiral CT images through the CT image processing unit, thereby avoiding the time wasted and segmentation errors caused by manual segmentation of the tumor lesion areas and normal organ areas, thereby improving the processing efficiency and accuracy of CT images. At the same time, when formulating radiotherapy plan information, the embodiment of the present invention utilizes a radiotherapy information determination device to perform a comprehensive analysis of the radiotherapy physician's radiotherapy demand information, the surgeon's surgical demand information, the tumor lesion area, the normal organ area, and the patient's characteristic information. This enables the formulated radiotherapy plan information to meet both the radiotherapy needs of the radiotherapy physician and the surgical needs of the surgeon during the operation, while also making the determined radiotherapy information suitable for the personalized physical sign information of the cancer patient, thereby enabling the formulated radiotherapy plan information to meet multiple needs.

[0097] Further, as Figure 1 In a specific implementation, an embodiment of the present invention provides a device for determining radiotherapy information, such as Figure 3 As shown, the device includes: an information acquisition unit 31, a CT image processing unit 32, and a radiotherapy information formulation unit 33.

[0098] The information acquisition unit 31 can be used to respond to the tumor radiotherapy area division signal of the liver cancer patient, obtain the multi-slice spiral CT image of the tumor of the liver cancer patient, the radiotherapy requirement information of the liver cancer patient at the radiotherapy terminal, the surgical requirement information of the liver cancer patient at the surgical terminal, and the patient characteristic information of the liver cancer patient.

[0099] The CT image processing unit 32 may be used to segment a tumor focus region and a normal organ region surrounding the tumor focus region from the multi-slice spiral CT image.

[0100] The radiotherapy information formulation unit 33 can be used to determine the radiotherapy plan information of the liver cancer patient based on the radiotherapy requirement information, the surgery requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information.

[0101] In a specific application scenario, in order to segment the tumor lesion area and the normal organ area from the multi-slice spiral CT image, such as Figure 4 As shown, the CT image processing unit 32 includes a processing module 321 and a first determining module 322 .

[0102] The processing module 321 may be used to perform grayscale processing on the multi-slice spiral CT image to obtain a grayscale image corresponding to the multi-slice spiral CT image.

[0103] The first determination module 322 can be used to determine a growth starting pixel point in the tumor lesion area in the grayscale image, determine a preset neighborhood of the growth starting pixel point, and select neighborhood pixels whose pixel value difference with the growth starting pixel point meets a preset difference condition within the preset neighborhood, and use the area formed by the neighborhood pixels and the growth starting pixel point as the growth area.

[0104] The first determination module 322 can be specifically used to use each pixel point in the growth area as a new growth starting pixel point, continue to determine a new preset neighborhood corresponding to the new growth starting pixel point, and select a new neighborhood pixel point in the new preset neighborhood whose pixel value difference with the new growth starting pixel point meets a preset difference condition, and use the area formed by the new neighborhood pixel point and the new growth starting pixel point as the new growth area, until no new neighborhood pixel point is found in the new preset neighborhood, and determine the finally determined new growth area as the tumor lesion area, and determine the area around the tumor lesion area as the normal organ area.

[0105] In a specific application scenario, in order to preprocess the CT image, the determination device further includes a division unit 34 , an addition unit 35 , an update unit 36 ​​, and a splicing unit 37 .

[0106] The CT image processing unit 32 may also be used to segment the grayscale image corresponding to the multi-slice spiral CT image into a plurality of non-overlapping sub-image blocks.

[0107] The division unit 34 can be used to determine the number of pixels of each gray level in each sub-image block, and divide the number of pixels of each gray level by the total number of pixels of the corresponding sub-image block to obtain the probability density of each gray level in each sub-image block.

[0108] The adding unit 35 can be used to sort the gray levels in ascending order, and take any gray level except the first gray level in the sorted gray levels as a current gray level, and add the probability density of each current gray level with the probability density of each gray level sorted before the current gray level to obtain the cumulative probability density corresponding to each current gray level.

[0109] The updating unit 36 ​​can be used to multiply the cumulative probability density with the maximum grayscale level in the corresponding sub-image block to obtain a new grayscale level corresponding to each of the current grayscale levels, and update each pixel value under each of the current grayscale levels to a pixel value corresponding to the new grayscale level to obtain each of the sub-image blocks after equalization.

[0110] The stitching unit 37 may be used to determine the original position of each of the sub-image blocks in the grayscale image, and stitch the equalized sub-image blocks according to the original position to obtain the pre-processed multi-slice spiral CT image.

[0111] In a specific application scenario, in order to determine radiotherapy plan information, the radiotherapy information formulation unit 33 includes: a trimming module 331 , a second determination module 332 , an acquisition module 333 , and an expansion module 334 .

[0112] The trimming module 331 can be used to trim the tumor lesion area and the normal organ area in the multi-slice spiral CT image based on the radiotherapy requirement information to obtain the trimmed tumor lesion area and the trimmed normal organ area.

[0113] The second determination module 332 may be configured to determine the portal vein cancer thrombus region in the multi-slice spiral CT image based on the trimmed tumor lesion region and the trimmed normal organ region.

[0114] The second determination module 332 can also be used to determine the lesion area to be removed and the prohibited radiotherapy irradiation area in the multi-slice spiral CT image based on the surgical requirement information, the trimmed tumor lesion area, the trimmed normal organ area, and the portal vein cancer thrombus area.

[0115] The second determination module 332 can also be used to determine the initial radiotherapy area of ​​the liver cancer patient in the multi-slice spiral CT image based on the trimmed tumor lesion area, the trimmed normal organ area, the venous cancer thrombus area, the lesion area to be resected, and the prohibited radiotherapy irradiation area.

[0116] The acquisition module 333 may be used to acquire patient positioning information when acquiring CT images of the liver cancer patient.

[0117] The expansion module 334 can be used to determine the expansion information for the initial radiotherapy area based on the patient positioning information, and expand the initial radiotherapy area based on the expansion information to obtain the radiotherapy area for the liver cancer patient.

[0118] The second determination module 332 may also be used to determine a radiotherapy dose for the radiotherapy area based on the area size of the radiotherapy area and the patient characteristic information.

[0119] In a specific application scenario, in order to formulate surgical treatment planning information, the determination device also includes: a three-dimensional reconstruction unit 38 and a fusion unit 39.

[0120] The information acquisition unit 31 may also be used to obtain target area and actual radiation dose distribution information of the liver cancer patient after actual radiotherapy in response to a radiotherapy end signal for the liver cancer patient based on the radiotherapy plan information.

[0121] The three-dimensional reconstruction unit 38 can be used to perform three-dimensional reconstruction on the multi-slice spiral CT image to obtain a three-dimensional image corresponding to the multi-slice spiral CT image.

[0122] The fusion unit 39 can be used to align and fuse the actual target area after radiotherapy, the actual radiation dose distribution information and the three-dimensional image to obtain the three-dimensional image with annotation information, wherein the annotation information includes the annotation box of the target area after radiotherapy and the actual radiation dose distribution information.

[0123] The radiotherapy information formulation unit 33 may also be used to determine surgical treatment planning information for the liver cancer patient based on the three-dimensional image with annotation information.

[0124] In a specific application scenario, the multi-slice spiral CT image includes: a late arterial phase CT image and a venous phase CT image of the liver cancer patient; in order to segment the tumor lesion area and the normal organ area from the CT image, the CT image processing unit 32 can be specifically used to input the late arterial phase CT image into a first image segmentation model constructed based on a deep learning architecture for image segmentation to obtain the hepatic artery area of ​​the liver cancer patient; and input the venous phase CT image into a second image segmentation model constructed based on a deep learning architecture for image segmentation to obtain the tumor lesion area, portal vein area, hepatic vein area, and bile duct area of ​​the liver cancer patient.

[0125] In a specific application scenario, in order to determine the radiotherapy plan information of a liver cancer patient, the radiotherapy information formulation unit 33 also includes: a cross module 335 and a prediction module 336 .

[0126] The second determination module 332 can also be used to respectively determine the radiotherapy feature vector corresponding to the radiotherapy requirement information, the surgical feature vector corresponding to the surgical requirement information, the tumor feature vector corresponding to the tumor lesion area, the organ feature vector corresponding to the normal organ area, and the patient feature vector corresponding to the patient feature information.

[0127] The cross module 335 may be used to perform cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a radiotherapy information cross vector.

[0128] The prediction module 336 may be used to input the radiotherapy information cross vector into a preset radiotherapy plan prediction model to perform radiotherapy plan prediction, thereby obtaining the radiotherapy plan information for the liver cancer patient.

[0129] In a specific application scenario, in order to perform cross processing on the vectors, the cross module 335 includes: a cross submodule and a transformation submodule.

[0130] The crossover submodule may be used to perform feature-level crossover processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a feature crossover vector.

[0131] The crossover submodule may also be used to perform element-level crossover processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain an element-level crossover vector.

[0132] The cross submodule may also be used to perform low-order cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a low-order cross vector.

[0133] The transformation submodule can be used to transform the characteristic cross vector, the element cross vector, and the low-order cross vector to obtain a radiotherapy information cross vector.

[0134] It should be noted that for other corresponding descriptions of the functional modules involved in the apparatus for determining radiotherapy information provided in an embodiment of the present invention, reference can be made to Figure 1 The corresponding description shown will not be repeated here.

[0135] Based on the above Figure 1As described above, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which implements the following steps when executed by a processor: in response to a tumor radiotherapy area division signal of a liver cancer patient, obtaining a multi-slice spiral CT image of the liver cancer patient's tumor, radiotherapy requirement information of the liver cancer patient at a radiotherapy terminal, surgical requirement information of the liver cancer patient at a surgical terminal, and patient characteristic information of the liver cancer patient; segmenting a tumor lesion area and a normal organ area surrounding the tumor lesion area from the multi-slice spiral CT image; and determining radiotherapy plan information of the liver cancer patient based on the radiotherapy requirement information, the surgical requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information.

[0136] Based on the above Figure 1 The process of determining radiotherapy information by the radiotherapy information determination device shown in the figure and the Figure 3 The embodiment of the device shown in the figure, the embodiment of the present invention also provides a physical structure diagram of a computer device, such as Figure 5 As shown, the computer device includes: a processor 41, a memory 42, and a computer program stored in the memory 42 and executable on the processor, wherein the memory 42 and the processor 41 are both arranged on a bus 43, and the processor 41 implements the following steps when executing the program: in response to a tumor radiotherapy area division signal of a liver cancer patient, obtaining a multi-slice spiral CT image of the liver cancer patient's tumor, radiotherapy requirement information of the liver cancer patient at a radiotherapy terminal, surgical requirement information of the liver cancer patient at a surgical terminal, and patient characteristic information of the liver cancer patient; segmenting a tumor lesion area and a normal organ area surrounding the tumor lesion area from the multi-slice spiral CT image; and determining the radiotherapy plan information of the liver cancer patient based on the radiotherapy requirement information, the surgical requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information.

[0137] Through the technical scheme of the present invention, the present invention constructs a device for determining radiotherapy information, obtains multi-slice spiral CT images of liver cancer patients, radiotherapy requirement information for liver cancer patients at a radiotherapy terminal, surgical requirement information for liver cancer patients at a surgical terminal, and patient characteristic information of liver cancer patients through an information acquisition unit in the determination device, segments tumor lesion areas and normal organ areas from multi-slice spiral CT images through a CT image processing unit in the determination device, and finally comprehensively analyzes the radiotherapy requirement information, surgical requirement information, tumor lesion areas, normal organ areas, and patient characteristic information through a radiotherapy information formulation unit in the determination device, and determines radiotherapy plan information based on the analysis results. Therefore, the present invention segments the tumor lesion areas and normal organ areas in multi-slice spiral CT images through the CT image processing unit, thereby avoiding the time wasted and segmentation errors caused by manual segmentation of tumor lesion areas and normal organ areas, thereby improving the processing efficiency and accuracy of CT images. At the same time, when formulating radiotherapy plan information, the embodiment of the present invention utilizes a radiotherapy information determination device to perform a comprehensive analysis of the radiotherapy physician's radiotherapy demand information, the surgeon's surgical demand information, the tumor lesion area, the normal organ area, and the patient's characteristic information. This enables the formulated radiotherapy plan information to meet both the radiotherapy needs of the radiotherapy physician and the surgical needs of the surgeon during the operation. At the same time, the determined radiotherapy information is suitable for the personalized physical sign information of the cancer patient, so that the formulated radiotherapy plan can further improve the surgical effect on the basis of improving the radiotherapy effect.

[0138] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0139] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A device for determining radiotherapy information, characterized in that: include: Information acquisition unit, CT image processing unit, radiotherapy information formulation unit; The information acquisition unit is used to respond to the tumor radiotherapy area division signal of the liver cancer patient, obtain the multi-slice spiral CT image of the liver cancer patient's tumor, the radiotherapy requirement information of the liver cancer patient from the radiotherapy terminal, the surgery requirement information of the liver cancer patient from the surgery terminal, and the patient characteristic information of the liver cancer patient; The CT image processing unit is used to segment the tumor lesion area and the normal organ area around the tumor lesion area from the multi-slice spiral CT image; The radiotherapy information formulation unit is used to determine the radiotherapy plan information of the liver cancer patient based on the radiotherapy requirement information, the surgery requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information.

2. The device according to claim 1, characterized in that The CT image processing unit includes a processing module and a first determining module; the CT image processing unit segments a tumor lesion area and a normal organ area around the tumor lesion area from the multi-slice spiral CT image, including: The processing module is used to perform grayscale processing on the multi-slice spiral CT image to obtain a grayscale image corresponding to the multi-slice spiral CT image; The first determination module is used to determine a growth starting pixel point in the tumor lesion area in the grayscale image, determine a preset neighborhood of the growth starting pixel point, and select neighborhood pixels whose pixel value difference with the growth starting pixel point satisfies a preset difference condition within the preset neighborhood, and use the area formed by the neighborhood pixels and the growth starting pixel point as the growth area; The first determination module is specifically used to use each pixel point in the growth area as a new growth starting pixel point, continue to determine a new preset neighborhood corresponding to the new growth starting pixel point, and select new neighborhood pixels in the new preset neighborhood whose pixel value difference with the new growth starting pixel point meets a preset difference condition, and use the area formed by the new neighborhood pixels and the new growth starting pixel point as the new growth area, until no new neighborhood pixels are found in the new preset neighborhood, and determine the finally determined new growth area as the tumor lesion area, and determine the area around the tumor lesion area as the normal organ area.

3. The device according to claim 1, characterized in that The invention also includes a division unit, an addition unit, an update unit, and a splicing unit; before the CT image processing unit segments the tumor lesion area and the normal organ area around the tumor lesion area from the multi-slice spiral CT image, the invention also includes: The CT image processing unit may also be used to segment the grayscale image corresponding to the multi-slice spiral CT image into a plurality of non-overlapping sub-image blocks; The division unit is used to determine the number of pixels of each gray level in each sub-image block, and divide the number of pixels of each gray level by the total number of pixels of the corresponding sub-image block to obtain the probability density of each gray level in each sub-image block; The adding unit is used to sort the gray levels in ascending order, and take any gray level except the first gray level in the sorted gray levels as a current gray level, and add the probability density of each current gray level with the probability density of each gray level sorted before the current gray level to obtain the cumulative probability density corresponding to each current gray level; The updating unit is used to multiply the cumulative probability density by the maximum gray level in the corresponding sub-image block to obtain a new gray level corresponding to each current gray level, and update each pixel value at each current gray level to a pixel value corresponding to the new gray level to obtain each sub-image block after equalization; The splicing unit is used to determine the original position of each of the sub-image blocks in the grayscale image, and splice the equalized sub-image blocks according to the original position to obtain the pre-processed multi-slice spiral CT image.

4. The device according to claim 1, characterized in that The radiotherapy information formulation unit includes: a trimming module, a second determination module, an acquisition module, and an expansion module; the radiotherapy information formulation unit determines the radiotherapy plan information of the liver cancer patient based on the radiotherapy requirement information, the surgery requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information, including: The trimming module is used to trim the tumor lesion area and the normal organ area in the multi-slice spiral CT image based on the radiotherapy requirement information to obtain the trimmed tumor lesion area and the trimmed normal organ area; The second determination module is used to determine the portal vein cancer thrombus area in the multi-slice spiral CT image based on the trimmed tumor lesion area and the trimmed normal organ area; The second determination module is further used to determine the lesion area to be removed and the prohibited radiotherapy irradiation area in the multi-slice spiral CT image based on the surgical requirement information, the trimmed tumor lesion area, the trimmed normal organ area, and the portal vein cancer thrombus area; The second determination module is further used to determine the initial radiotherapy area of ​​the liver cancer patient in the multi-slice spiral CT image based on the trimmed tumor lesion area, the trimmed normal organ area, the venous cancer thrombus area, the lesion area to be resected, and the radiotherapy irradiation prohibited area; The acquisition module is used to acquire patient positioning information when acquiring CT images of the liver cancer patient; The expansion module is used to determine the expansion information for the initial radiotherapy area based on the patient positioning information, and expand the initial radiotherapy area based on the expansion information to obtain the radiotherapy area for the liver cancer patient; The second determination module is further used to determine the radiotherapy dose for the radiotherapy area based on the area size of the radiotherapy area and the patient characteristic information.

5. The device according to claim 1, characterized in that Also includes: 3D reconstruction unit, fusion unit; After the radiotherapy information formulation unit determines the radiotherapy plan information of the liver cancer patient based on the radiotherapy requirement information, the surgery requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information, the further comprising: The information acquisition unit is further used to obtain target area and actual radiation dose distribution information of the liver cancer patient after actual radiotherapy in response to a radiotherapy end signal of the liver cancer patient based on the radiotherapy plan information; The three-dimensional reconstruction unit is used to perform three-dimensional reconstruction on the multi-slice spiral CT image to obtain a three-dimensional image corresponding to the multi-slice spiral CT image; The fusion unit is used to align and fuse the target area after the actual radiotherapy, the actual radiation dose distribution information and the three-dimensional image to obtain the three-dimensional image with annotation information, wherein the annotation information includes a annotation frame of the target area after the actual radiotherapy and the actual radiation dose distribution information; The radiotherapy information formulation unit is further used to determine the surgical treatment planning information of the liver cancer patient based on the three-dimensional image with the annotation information.

6. The device according to claim 1, characterized in that The multi-slice spiral CT images include: late arterial phase CT images and venous phase CT images of the liver cancer patient; The CT image processing unit segments a tumor lesion region and a normal organ region surrounding the tumor lesion region from the multi-slice spiral CT image, comprising: The CT image processing unit is used to input the late arterial CT image into a first image segmentation model constructed based on a deep learning architecture to perform image segmentation, so as to obtain the hepatic artery region of the liver cancer patient; The CT image processing unit is used to input the venous phase CT image into a second image segmentation model constructed based on a deep learning architecture to perform image segmentation to obtain the tumor lesion area, portal vein area, hepatic vein area, and bile duct area of ​​the liver cancer patient.

7. The device according to claim 1, characterized in that The radiotherapy information formulation unit further includes: a cross module and a prediction module; the radiotherapy information formulation unit determines the radiotherapy plan information of the liver cancer patient based on the radiotherapy requirement information, the surgery requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information, including: The second determination module is further used to respectively determine the radiotherapy feature vector corresponding to the radiotherapy requirement information, the surgery feature vector corresponding to the surgery requirement information, the tumor feature vector corresponding to the tumor lesion area, the organ feature vector corresponding to the normal organ area, and the patient feature vector corresponding to the patient feature information; The cross module is used to perform cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a radiotherapy information cross vector; The prediction module is used to input the radiotherapy information cross vector into a preset radiotherapy plan prediction model to perform radiotherapy plan prediction, so as to obtain the radiotherapy plan information of the liver cancer patient.

8. The device according to claim 7, characterized in that The cross module includes: a cross submodule and a transformation submodule; the cross module performs cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a radiotherapy information cross vector, including: The cross submodule is used to perform feature-level cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a feature cross vector; The crossover submodule is further used to perform element-level crossover processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain an element crossover vector; The cross submodule is further used to perform low-order cross processing on the radiotherapy feature vector, the surgery feature vector, the tumor feature vector, the organ feature vector, and the patient feature vector to obtain a low-order cross vector; The transformation submodule is used to transform the characteristic cross vector, the element cross vector, and the low-order cross vector to obtain a radiotherapy information cross vector.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the following steps are implemented: In response to a tumor radiotherapy area division signal of a liver cancer patient, obtaining a multi-slice spiral CT image of the liver cancer patient's tumor, radiotherapy requirement information of the liver cancer patient at a radiotherapy terminal, surgical requirement information of the liver cancer patient at a surgical terminal, and patient characteristic information of the liver cancer patient; Segmenting a tumor lesion region and a normal organ region surrounding the tumor lesion region from the multi-slice spiral CT image; Based on the radiotherapy requirement information, the surgical requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information, the radiotherapy plan information of the liver cancer patient is determined.

10. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the computer program is executed by a processor, the following steps are implemented: In response to a tumor radiotherapy area division signal of a liver cancer patient, obtaining a multi-slice spiral CT image of the liver cancer patient's tumor, radiotherapy requirement information of the liver cancer patient at a radiotherapy terminal, surgical requirement information of the liver cancer patient at a surgical terminal, and patient characteristic information of the liver cancer patient; Segmenting a tumor lesion region and a normal organ region surrounding the tumor lesion region from the multi-slice spiral CT image; Based on the radiotherapy requirement information, the surgical requirement information, the tumor lesion area, the normal organ area, and the patient characteristic information, the radiotherapy plan information of the liver cancer patient is determined.

Images