A method, device and program product for predicting the location of the target area of ​​the lung abdomen under CT guidance

Abstract

The present application relates to the field of intelligent medical treatment, and specifically to a method, device and program product for predicting the position of the target area of ​​the lung and abdomen under CT guidance. It includes S1, obtaining the first CT image and respiratory motion data of the target area of ​​the lung and abdomen; S2, extracting the position of the first CT image of the target area of ​​the lung and abdomen; S3, inputting the position of the first CT image of the target area of ​​the lung and abdomen and the respiratory motion data into a mathematical model for prediction to obtain the moving direction and moving distance of the target area of ​​the lung and abdomen; S4, obtaining a second CT image based on the moving direction and moving distance of the target area of ​​the lung and abdomen; wherein the mathematical model is obtained by calculating the relationship between the movement changes of the CT images of the target area of ​​the lung and abdomen at different time points and the respiratory motion data on a time axis. The present application can predict and measure in real time the position changes of the target area of ​​the lung and abdomen in the up and down directions on the time axis with respiratory motion, and has great clinical value.

Description

Technical Field

[0001] The present application relates to the field of intelligent medical treatment, and specifically to a method, device and program product for predicting the location of a target area in the lung abdomen under CT guidance. Background Art

[0002] At present, lung cancer has become the leading cause of cancer-related death in my country and around the world. The clinical treatment principle for lung cancer is: early diagnosis and early treatment. With the increasing popularity of chest CT in clinical applications, more and more small lung nodules suspected of early lung cancer are discovered early. During the follow-up of these small nodules, when the volume of ground-glass nodules increases or the solid components increase, and the solid part of some solid nodules exceeds 6mm, clinical intervention is required. Minimally invasive video-assisted thoracic surgery (VATS) is the preferred method for treating these small nodules. However, studies have shown that a high proportion of thoracoscopic resection of small lung nodules cannot be accurately located because they cannot be seen or palpated during surgery, and traditional thoracotomy is performed instead; when the diameter of the lung nodule is less than 10mm or the distance from the pleura is greater than 10mm, it is difficult to accurately locate the small lung nodules during surgery, especially pure ground-glass nodules. Preoperative CT-guided micro-coil positioning of small lung nodules can effectively solve this problem, so that small nodules in the lung can be accurately removed.

[0003] However, due to the influence of respiratory movement, there are difficulties in the puncture of small lung nodules under CT guidance. Lung nodules will shift with respiratory movement, especially nodules in the lower lobe, which can shift up to 50mm in the up and down direction with respiratory movement. Many studies have shown that due to the influence of respiratory movement, nodules located in the lower lobe are risk factors for failure of CT-guided lung puncture and risk factors for complications such as pneumothorax. Although some CT-guided puncture navigation devices have been launched, the targets set by these navigation devices are fixed targets, and the problem of lung nodules shifting with breathing during puncture has not been solved. Puncture under the guidance of C-arm cone beam CT or CT fluoroscopy can dynamically observe the displacement of lung nodules with breathing in real time, but these methods will cause ionizing radiation damage to the operator. Summary of the invention

[0004] In view of the above problems, the present invention proposes a method for predicting the location of the target area of ​​the lung abdomen under CT guidance, which specifically includes:

[0005] S1, obtaining the first CT image and respiratory motion data of the target area of ​​the lung abdomen;

[0006] S2, extracting the position of the first CT image of the lung abdomen target area based on the first CT image of the lung abdomen target area;

[0007] S3, inputting the position of the first CT image of the target area of ​​the lung abdomen and the respiratory movement data into a mathematical model to predict the moving direction and moving distance of the target area of ​​the lung abdomen;

[0008] S4, obtaining a second CT image based on the moving direction and moving distance of the target area of ​​the lung abdomen;

[0009] The mathematical model is obtained by calculating the relationship between the moving direction and moving distance of the CT image of the target area of ​​the lung abdomen at different time points and the respiratory motion data on a time axis.

[0010] Furthermore, the steps for constructing the mathematical model are as follows:

[0011] Obtain N CT images and respiratory motion data of the patient's lung and abdominal target area at different time points, where N is a natural number greater than or equal to 5;

[0012] Obtaining the position of the lung and abdomen target area based on the N CT images of the lung and abdomen target area at different time points;

[0013] Calculating the moving distance and moving direction of the lung abdomen target area based on the position of the lung abdomen target area;

[0014] A mathematical model is obtained by calculating a mathematical relationship between the moving distance and the moving direction of the target area of ​​the lung abdomen and the respiratory motion data;

[0015] Optionally, the moving distance calculation of the lung abdomen target area is based on the position of the lung abdomen target area in the first CT image, and the relative moving distance and moving direction of the lung abdomen target area in other images are calculated;

[0016] Optionally, the respiratory motion data is chest and abdomen respiratory motion waveform data;

[0017] Optionally, the respiratory motion data is real-time respiratory motion data, and the position of the first CT image of the lung abdominal target area and the real-time respiratory motion data are input into a mathematical model for prediction to obtain the real-time moving direction and real-time moving distance of the lung abdominal target area, and a real-time second CT image is obtained based on the real-time moving direction and real-time moving distance of the lung abdominal target area.

[0018] The method step S4 is specifically as follows: the position of the first CT image of the lung abdominal target area is calculated with the moving direction and moving distance of the lung abdominal target area to obtain the position of the second CT image of the lung abdominal target area, and image reconstruction is performed based on the position of the second CT image of the lung abdominal target area to obtain the second CT image.

[0019] The S3 is replaced by S31: inputting the first CT image and the respiratory motion data of the target area of ​​the lung abdomen into the absolute mathematical model to obtain a second CT image; wherein the absolute mathematical model is obtained by calculating the relationship between the CT image and the respiratory motion data at different time points on a time axis;

[0020] Optionally, the steps of constructing the absolute mathematical model are specifically:

[0021] Obtain N CT images and respiratory motion data of the patient's lung and abdominal target area at different time points, where N is a natural number greater than or equal to 5;

[0022] Based on the mathematical relationship calculation of N CT images of the target area of ​​the lung abdomen at different time points and the respiratory motion data, an absolute mathematical model is obtained;

[0023] Optionally, the mathematical relationship is a mapping relationship.

[0024] Replace S3 with S32:

[0025] The first CT image of the target area of ​​the lung abdomen and the respiratory movement data are input into the prediction model to predict the moving direction and moving distance of the target area of ​​the lung abdomen.

[0026] Optionally, the steps of constructing the prediction model are specifically as follows:

[0027] Obtain CT image sets and labels of the target area of ​​the lung abdomen at different time points, as well as respiratory motion data and labels;

[0028] The CT image set and labels of the target area of ​​the lung abdomen at different time points, the respiratory motion data and labels are respectively input into the neural network for feature extraction to obtain the image feature vector and the respiratory motion feature vector;

[0029] Performing feature fusion on the image feature vector and the respiratory motion feature vector to obtain a fusion feature;

[0030] The moving direction and moving distance of the target area of ​​the lung abdomen are predicted based on the fusion features.

[0031] The method further includes obtaining clinical physiological and biochemical data, inputting the clinical physiological and biochemical data, CT images of the target area of ​​the lung and abdomen, and respiratory movement data into a second prediction model to predict the moving direction and moving distance of the target area of ​​the lung and abdomen;

[0032] Optionally, the clinical physiological and biochemical data include one or more of the following: gender, age, height, weight, chest circumference, abdominal circumference, body position and vital capacity.

[0033] Optionally, the specific steps of the method for constructing the second prediction model are:

[0034] Obtain clinical physiological and biochemical data sets and labels, CT image sets and labels of the target area of ​​the lung abdomen at different time points, and respiratory motion data and labels;

[0035] The clinical physiological and biochemical data set and labels, the CT image set and labels of the target area of ​​the lung abdomen at different time points, and the respiratory motion data and labels are respectively input into the neural network for feature extraction to obtain the clinical physiological and biochemical feature vector, the image feature vector, and the respiratory motion feature vector;

[0036] Performing feature fusion on the clinical physiological and biochemical feature vector, the imaging feature vector, and the respiratory motion feature vector to obtain a fusion feature;

[0037] The moving direction and moving distance of the target area of ​​the lung abdomen are predicted based on the fusion features.

[0038] The feature fusion is performed through a neural network to obtain a fusion feature;

[0039] Optionally, the neural network is any one or more of the following: Transformer, CNN, Attention, ResNet, LSTM, GAT;

[0040] Optionally, the neural network performs feature extraction through Transformer to obtain a feature vector.

[0041] The object of the present invention is to provide a computer program product having a computer program or instructions thereon, wherein the method is executed by a processor to implement the above-mentioned method for predicting the position of the target area of ​​the lung abdomen under CT guidance.

[0042] The object of the present invention is to provide a computer device, comprising a memory, a processor, and a computer program or instruction stored in the memory, including:

[0043] The computer program or instructions are executed by the processor to implement the above-mentioned method for predicting the position of the target area in the abdomen of the lung under CT guidance.

[0044] The object of the present invention is to provide a computer-readable storage medium having a computer program or instruction stored thereon, including: the computer program or instruction is executed by a processor to perform the above-mentioned method for predicting the position of a target area of ​​the lung abdomen under CT guidance.

[0045] Advantages of the present invention:

[0046] The relevant parameters of respiratory movement on the timeline and the parameters of the vertical changes in the lung and abdominal target areas at different positions in the CT image are associated using mathematical algorithms to establish an innovative mathematical model. This mathematical model can predict the vertical change trajectory of the lung and abdominal target areas at different positions on the timeline, and can assist CT puncture in actual clinical applications, reduce surgical difficulty, and improve the success rate of surgery.

[0047] By mathematically mapping the CT images at different time nodes with the respiratory motion data, a mathematical model is obtained, which can be used to directly predict the CT image at the next time node. In addition, real-time images can be predicted through real-time respiratory motion data, which helps to choose the appropriate time for needle insertion or adjust the appropriate angle for needle insertion.

[0048] A prediction model is proposed. Based on the feature learning of CT images of the target area in the abdomen of the lung and the respiratory motion data, the relationship between the target area in the abdomen of the lung and the respiratory motion data in the image is obtained. This model is used to guide CT-guided puncture. By real-time monitoring of changes in patients' respiratory motion-related parameters during surgery, the changing trajectory of small lung nodules in the upper and lower directions can be predicted in real time, and the appropriate time for needle insertion or the appropriate angle of needle insertion can be selected to reduce the number of CT scans during surgery (reduce radiation dose) and reduce complications. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0050] Figure 1 A schematic flow chart of a method for predicting the location of a target area in the lung abdomen under CT guidance provided by an embodiment of the present invention;

[0051] Figure 2 A schematic diagram of a system for predicting the location of a target area in the abdomen of the lung under CT guidance provided by an embodiment of the present invention;

[0052] Figure 3 A schematic diagram of a device for predicting the location of a target area in the lung abdomen under CT guidance provided by an embodiment of the present invention;

[0053] Figure 4 A structural diagram of a prediction model provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0054] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0055] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The sequence numbers of the operations, such as S101, S102, etc., are only used to distinguish between different operations, and the sequence numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0056] Figure 1 A schematic diagram of a method for predicting the location of a target area of ​​the lung abdomen under CT guidance provided by an embodiment of the present invention specifically includes:

[0057] S1: Acquire the first CT image and respiratory motion data of the target area of ​​the lung abdomen;

[0058] In one embodiment, the respiratory motion data is chest and abdomen respiratory motion waveform data.

[0059] In one embodiment, the respiratory motion data is real-time respiratory motion data.

[0060] In a specific embodiment, respiratory movement-related parameters (chest and abdominal mobility) are simultaneously collected using a respiratory monitoring device when the patient undergoes CT-guided puncture. During a complete CT-guided puncture, multiple CT scans are performed, usually about 5 times. The position of the lung nodule in the image obtained by the first CT scan is used as a reference, and the position of the nodule in the CT images obtained at different time points thereafter is compared with the first position to obtain the displacement distance (mm) in the up and down direction.

[0061] In a specific embodiment, a total of 150 patients were included in the study, and the data of the first 120 patients (including small nodules located in different lung segments) were used for data matching modeling to establish a prediction model based on changes in chest and abdominal activity. The data of the last 30 patients were used for verification.

[0062] S2: extracting the position of the first CT image of the lung abdomen target area based on the first CT image of the lung abdomen target area;

[0063] In one embodiment, target detection is performed on a first CT image of a target area in the lung abdomen to obtain a position of the target area;

[0064] In one embodiment, the target detection obtains the position of the target area through a target detection network; the target detection network is one or more of the following: R-CNN, Fast R-CNN, YOLO, SSD, Mask R-CNN, RetinaNet.

[0065] S3: inputting the position of the first CT image of the target area of ​​the lung abdomen and the respiratory movement data into a mathematical model to predict the moving direction and moving distance of the target area of ​​the lung abdomen;

[0066] The mathematical model is obtained by calculating the relationship between the moving direction and moving distance of the CT image of the target area of ​​the lung abdomen at different time points and the respiratory motion data on a time axis.

[0067] In one embodiment, the steps of constructing the mathematical model are specifically as follows:

[0068] Obtain N CT images and respiratory motion data of the patient's lung and abdominal target area at different time points, where N is a natural number greater than or equal to 5;

[0069] Obtaining the position of the lung and abdomen target area based on the N CT images of the lung and abdomen target area at different time points;

[0070] Calculating the moving distance and moving direction of the lung abdomen target area based on the position of the lung abdomen target area;

[0071] The mathematical model is obtained by calculating the mathematical relationship between the moving distance and the moving direction of the target area in the abdomen of the lung and the respiratory motion data.

[0072] Optionally, the moving distance calculation of the lung abdomen target area is based on the position of the lung abdomen target area in the first CT image, and the relative moving distance and moving direction of the lung abdomen target area in other images are calculated.

[0073] In another embodiment, S3 is replaced by S31: the first CT image and respiratory motion data of the target area of ​​the lung abdomen are input into an absolute mathematical model to obtain a second CT image; wherein the absolute mathematical model is obtained by calculating the relationship between the CT image and the respiratory motion data at different time points on a time axis.

[0074] In one embodiment, the steps of constructing the absolute mathematical model are specifically as follows:

[0075] Obtain N CT images and respiratory motion data of the patient's lung and abdominal target area at different time points, where N is a natural number greater than or equal to 5;

[0076] Based on the mathematical relationship calculation of N CT images of the target area of ​​the lung abdomen at different time points and the respiratory motion data, an absolute mathematical model is obtained;

[0077] Optionally, the mathematical relationship is a mapping relationship.

[0078] In another embodiment, the S3 is replaced by S32:

[0079] The first CT image of the target area of ​​the lung abdomen and the respiratory movement data are input into the prediction model to predict the moving direction and moving distance of the target area of ​​the lung abdomen.

[0080] In one embodiment, the steps of constructing the prediction model are specifically as follows:

[0081] Obtain CT image sets and labels of the target area of ​​the lung abdomen at different time points, as well as respiratory motion data and labels;

[0082] The CT image set and labels of the target area of ​​the lung abdomen at different time points, the respiratory motion data and labels are respectively input into the neural network for feature extraction to obtain the image feature vector and the respiratory motion feature vector;

[0083] Performing feature fusion on the image feature vector and the respiratory motion feature vector to obtain a fusion feature;

[0084] The moving direction and moving distance of the target area of ​​the lung abdomen are predicted based on the fusion features.

[0085] In another embodiment, the method further includes obtaining clinical physiological and biochemical data, inputting the clinical physiological and biochemical data, CT images of the target area of ​​the lung and abdomen, and respiratory motion data into a second prediction model to predict the moving direction and moving distance of the target area of ​​the lung and abdomen;

[0086] Optionally, the clinical physiological and biochemical data include one or more of the following: gender, age, height, weight, chest circumference, abdominal circumference, body position and vital capacity.

[0087] In one embodiment, the method for constructing the second prediction model comprises the following specific steps:

[0088] Obtain clinical physiological and biochemical data sets and labels, CT image sets and labels of the target area of ​​the lung abdomen at different time points, and respiratory motion data and labels;

[0089] The clinical physiological and biochemical data set and labels, the CT image set and labels of the target area of ​​the lung abdomen at different time points, and the respiratory motion data and labels are respectively input into the neural network for feature extraction to obtain the clinical physiological and biochemical feature vector, the image feature vector, and the respiratory motion feature vector;

[0090] Performing feature fusion on the clinical physiological and biochemical feature vector, the imaging feature vector, and the respiratory motion feature vector to obtain a fusion feature;

[0091] The moving direction and moving distance of the target area of ​​the lung abdomen are predicted based on the fusion features.

[0092] In one embodiment, the feature fusion is performed through a neural network to obtain a fusion feature.

[0093] In one embodiment, the neural network is any one or more of the following: Transformer, CNN, Attention, ResNet, LSTM, GAT.

[0094] In one embodiment, the neural network performs feature extraction through Transformer to obtain a feature vector.

[0095] In a specific embodiment, when the patient undergoes CT-guided puncture, information on relevant covariates is collected: gender, age, height, weight, chest circumference, abdominal circumference, body position, and vital capacity.

[0096] In a specific embodiment, the displacement distance of the pulmonary nodules on the CT scans at different time points is matched with the respiratory motion-related parameters on the time axis, and a mathematical algorithm is used to establish an innovative mathematical model to associate the respiratory motion parameters with the displacement distance of the pulmonary nodules in the up and down directions. After the respiratory motion parameters on the time axis and the position of the pulmonary nodules in the first CT image are input into the mathematical model, the position changes of the pulmonary nodules in the up and down directions on the time axis with the respiratory motion can be predicted in real time.

[0097] In a specific embodiment, a deep learning algorithm is used for modeling, and a deep artificial neural network is used to construct an implicit mathematical model to establish a prediction model based on changes in chest and abdominal activity. In this study, the input information of the prediction model based on changes in chest and abdominal activity includes: chest and abdominal respiratory movement waveforms, clinical physiological and biochemical covariate information, and CT image information. In order to fully explore the interactive relationship between information of different dimensions, this study intends to use the Transformer algorithm as the backbone network to extract feature vectors of information of different dimensions respectively, and then use the Transformer algorithm's attention-based encoder to fuse features of different dimensions, explore the correlation between features, and finally use multiple fully connected layers to output the change direction and distance of lung nodule movement. In addition, the present invention intends to use a stochastic gradient descent strategy, a cross entropy loss function, and a mean square error loss function to train and optimize the model. This deep learning model is also suitable for the construction of this model, and its structure is as follows Figure 4 shown.

[0098] S4: obtaining a second CT image based on the moving direction and moving distance of the target area of ​​the lung abdomen;

[0099] In one embodiment, the respiratory motion data is real-time respiratory motion data. The position of the first CT image of the lung abdominal target area and the real-time respiratory motion data are input into a mathematical model for prediction to obtain the real-time moving direction and real-time moving distance of the lung abdominal target area, and a real-time second CT image is obtained based on the real-time moving direction and real-time moving distance of the lung abdominal target area.

[0100] In one embodiment, step S4 of the method is specifically as follows: the position of the first CT image of the lung abdominal target area is calculated together with the moving direction and moving distance of the lung abdominal target area to obtain the position of the second CT image of the lung abdominal target area, and image reconstruction is performed based on the position of the second CT image of the lung abdominal target area to obtain the second CT image.

[0101] In a specific embodiment, the respiratory motion change parameters on the model time axis, the position of the lung nodules in the CT scan at the first time point and the related covariates are first given, and then the time points of the subsequent CT scans are given, so that the model is predicted, the position change of the lung nodules on the CT images at the subsequent time points relative to the first scan in the up and down direction, and the results obtained by the model are compared with the actual changes of the nodules in the up and down directions in the subsequent CT scan images. It is of great clinical significance to establish a non-invasive and radiation-free method to monitor and predict the real-time trajectory of the displacement of lung nodules with respiratory motion to guide CT-guided lung puncture, which can improve the success rate of surgery, reduce the radiation dose of patients, shorten the operation time and reduce complications. It can solve the problem of lung nodules being affected by respiratory motion during puncture, and can be applied to various CT-guided lung puncture surgeries. In addition to CT-guided lung nodule puncture positioning before VATS surgery, it can also be applied to CT-guided percutaneous lung biopsy, tumor ablation, and radioactive particle implantation. In addition, this solution can also be introduced into CT-guided puncture navigation equipment to upgrade it and make it more intelligent.

[0102] The embodiments of the present disclosure also provide a computer program product or system, including a computer program, which, when executed by a processor, implements the steps of the method for predicting the position of the target area of ​​the lung abdomen under CT guidance.

[0103] In one embodiment, a first CT image of a lung-abdomen target area and respiratory motion data are acquired, the position of the first CT image of the lung-abdomen target area is extracted based on the first CT image of the lung-abdomen target area, the position of the first CT image of the lung-abdomen target area and the respiratory motion data are input into a mathematical model for prediction to obtain a moving direction and a moving distance of the lung-abdomen target area, the position of the first CT image of the lung-abdomen target area and the moving direction and the moving distance of the lung-abdomen target area are calculated to obtain a position of a second CT image of the lung-abdomen target area, and image reconstruction is performed based on the position of the second CT image of the lung-abdomen target area to obtain a second CT image.

[0104] In one embodiment, a first CT image of a lung and abdomen target area and real-time respiratory motion data are acquired, the position of the first CT image of the lung and abdomen target area is extracted based on the first CT image of the lung and abdomen target area, the position of the first CT image of the lung and abdomen target area and the real-time respiratory motion data are input into a mathematical model for prediction to obtain a moving direction and a moving distance of the lung and abdomen target area, the position of the first CT image of the lung and abdomen target area and the moving direction and the moving distance of the lung and abdomen target area are calculated to obtain a position of a second real-time CT image of the lung and abdomen target area, and image reconstruction is performed based on the position of the second real-time CT image of the lung and abdomen target area to obtain a second real-time CT image.

[0105] In one embodiment, clinical physiological and biochemical data, a first CT image of the lung and abdomen target area, and respiratory motion data are obtained, and the clinical physiological and biochemical data, the CT image of the lung and abdomen target area, and the respiratory motion data are input into a second prediction model for prediction to obtain a moving direction and a moving distance of the lung and abdomen target area, the position of the first CT image of the lung and abdomen target area is calculated with the moving direction and the moving distance of the lung and abdomen target area to obtain the position of the second CT image of the lung and abdomen target area, and image reconstruction is performed based on the position of the second CT image of the lung and abdomen target area to obtain a second CT image.

[0106] Figure 2 A schematic diagram of a system for predicting the location of a target area of ​​the lung abdomen under CT guidance provided by an embodiment of the present invention specifically includes:

[0107] Figure 3 A schematic diagram of a device for predicting the location of a target area of ​​the lung abdomen under CT guidance provided by an embodiment of the present invention specifically includes:

[0108] A memory and a processor; the memory is used to store program instructions; the processor is used to call the program instructions, and when the program instructions are executed, any one of the above-mentioned methods for predicting the position of the target area of ​​the lung abdomen under CT guidance is performed.

[0109] The embodiment of the present disclosure further provides a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, is any one of the above-mentioned methods for predicting the position of a target area in the abdomen of the lung under CT guidance.

[0110] The verification results of this verification embodiment show that assigning inherent weights to indications can improve the performance of the method relative to the default setting. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the coupling or direct coupling or communication connection between each other shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms. The unit described as a separate component may or may not be physically separated, and the component displayed as a unit may or may not be a physical unit, that is, it may be located in one place, or it may be distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units. A person of ordinary skill in the art may understand that all or part of the steps in the various methods of the above-mentioned embodiments may be completed by instructing the relevant hardware through a program, and the program may be stored in a computer-readable storage medium, and the storage medium may include: a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.

[0111] A person skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be implemented by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. The above-mentioned medium storage can be a read-only memory, a disk or an optical disk, etc.

[0112] The above is a detailed introduction to a computer device provided by the present invention. For a person skilled in the art, according to the concept of the embodiments of the present invention, there may be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for predicting the location of a target area of ​​the lung abdomen under CT guidance, characterized in that: include: S1, obtaining the first CT image and respiratory motion data of the target area of ​​the lung abdomen; S2, extracting the position of the first CT image of the lung abdomen target area based on the first CT image of the lung abdomen target area; S3, inputting the position of the first CT image of the lung abdomen target area and the respiratory motion data into a mathematical model to predict the moving direction and moving distance of the lung abdomen target area; the respiratory motion data is real-time respiratory motion data, and the position of the first CT image of the lung abdomen target area and the real-time respiratory motion data are input into the mathematical model to predict the real-time moving direction and real-time moving distance of the lung abdomen target area; S4, obtaining a real-time second CT image based on the real-time moving direction and real-time moving distance of the target area of ​​the lung abdomen; specifically: calculating the position of the first CT image of the target area of ​​the lung abdomen with the moving direction and moving distance of the target area of ​​the lung abdomen to obtain the position of the second CT image of the target area of ​​the lung abdomen, and reconstructing the image based on the position of the second CT image of the target area of ​​the lung abdomen to obtain the second CT image; The mathematical model is obtained by calculating the relationship between the moving direction and moving distance of the CT image of the target area of ​​the lung abdomen at different time points and the respiratory motion data on a time axis.

2. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 1, characterized in that: The specific steps of constructing the mathematical model are: Obtain N CT images and respiratory motion data of the patient's lung and abdominal target area at different time points, where N is a natural number greater than or equal to 5; Obtaining the position of the lung and abdomen target area based on the N CT images of the lung and abdomen target area at different time points; Calculating the moving distance and moving direction of the lung abdomen target area based on the position of the lung abdomen target area; The mathematical model is obtained by calculating the mathematical relationship between the moving distance and the moving direction of the target area in the abdomen of the lung and the respiratory motion data.

3. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 1, characterized in that: The moving distance calculation of the lung abdomen target area is based on the position of the lung abdomen target area in the first CT image, and the relative moving distance and moving direction of the lung abdomen target area in other images are calculated.

4. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 1, characterized in that: The respiratory motion data are chest and abdomen respiratory motion waveform data.

5. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 1, characterized in that: The S3 is replaced by S31: inputting the first CT image and respiratory motion data of the target area of ​​the lung abdomen into the absolute mathematical model to obtain a second CT image; wherein the absolute mathematical model is obtained by calculating the relationship between the CT image and the respiratory motion data at different time points on a time axis.

6. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 5, characterized in that: The construction steps of the absolute mathematical model are specifically as follows: Obtain N CT images and respiratory motion data of the patient's lung and abdominal target area at different time points, where N is a natural number greater than or equal to 5; The absolute mathematical model is obtained by calculating the mathematical relationship between N CT images of the target area of ​​the lung abdomen at different time points and the respiratory motion data.

7. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 2 or 6, characterized in that: The mathematical relationship is a mapping relationship.

8. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 1, characterized in that: Replace S3 with S32: The first CT image of the target area of ​​the lung abdomen and the respiratory movement data are input into the prediction model to predict the moving direction and moving distance of the target area of ​​the lung abdomen.

9. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 8, characterized in that: The steps for constructing the prediction model are specifically as follows: Obtain CT image sets and labels of the target area of ​​the lung abdomen at different time points, as well as respiratory motion data and labels; The CT image set and labels of the target area of ​​the lung abdomen at different time points, the respiratory motion data and labels are respectively input into the neural network for feature extraction to obtain the image feature vector and the respiratory motion feature vector; Performing feature fusion on the image feature vector and the respiratory motion feature vector to obtain a fusion feature; The moving direction and moving distance of the target area of ​​the lung abdomen are predicted based on the fusion features.

10. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 8, characterized in that: The method also includes obtaining clinical physiological and biochemical data, inputting the clinical physiological and biochemical data, CT images of the target area of ​​the lung and abdomen, and respiratory movement data into a second prediction model to predict the moving direction and distance of the target area of ​​the lung and abdomen.

11. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 10, characterized in that: The clinical physiological and biochemical data include one or more of the following: gender, age, height, weight, chest circumference, abdominal circumference, body position and vital capacity.

12. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 10, characterized in that: The specific steps of the method for constructing the second prediction model are: Obtain clinical physiological and biochemical data sets and labels, CT image sets and labels of the target area of ​​the lung abdomen at different time points, and respiratory motion data and labels; The clinical physiological and biochemical data set and labels, the CT image set and labels of the target area of ​​the lung abdomen at different time points, and the respiratory motion data and labels are respectively input into the neural network for feature extraction to obtain the clinical physiological and biochemical feature vector, the image feature vector, and the respiratory motion feature vector; Performing feature fusion on the clinical physiological and biochemical feature vector, the imaging feature vector, and the respiratory motion feature vector to obtain a fusion feature; The moving direction and moving distance of the target area of ​​the lung abdomen are predicted based on the fusion features.

13. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 9 or 12, characterized in that: The feature fusion is performed through a neural network to obtain a fusion feature.

14. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 13, characterized in that: The neural network is any one or more of the following: Transformer, CNN, Attention, ResNet, LSTM, GAT.

15. The method for predicting the location of the target area of ​​the lung abdomen under CT guidance according to claim 13, characterized in that: The neural network performs feature extraction through Transformer to obtain a feature vector.

16. A computer program product having a computer program or instructions thereon, characterized in that: The method is executed by a processor to implement the method for predicting the position of the target area in the abdomen of the lung under CT guidance as described in any one of claims 1-15.

17. A computer device comprising a memory, a processor and a computer program or instruction stored in the memory, characterized in that: include: The computer program or instructions are executed by a processor to implement the method for predicting the position of the target area in the abdomen of the lung under CT guidance as described in any one of claims 1-15.

18. A computer-readable storage medium having a computer program or instruction stored thereon, comprising: The computer program or instructions are executed by a processor to implement the method for predicting the position of the target area in the abdomen of the lung under CT guidance as described in any one of claims 1-15.

Images