Kidney operation auxiliary path planning system based on artificial intelligence

Abstract

The invention discloses a kidney operation auxiliary path planning system based on artificial intelligence, and relates to the technical field of medical operation assistance. Comprising an intelligent image recognition and processing module, a personalized surgical path planning module, a real-time surgical navigation and adjustment module, an integrated medical database and knowledge base, a data analysis and machine learning module and a user-friendly interaction interface. By integrating the deep learning model, key anatomical structures and lesion areas such as kidneys, blood vessels and tumors can be automatically and accurately recognized, and an accurate basis is provided for planning of an operation path; the precise recognition capability obviously improves the precision and safety of the operation, and reduces the accidental injury and complication risk in the operation process.

Description

Technical Field

[0001] The present invention relates to the field of medical surgery auxiliary technology, and in particular to an artificial intelligence-based kidney surgery auxiliary path planning system. Background Art

[0002] In the modern medical field, surgical path planning is one of the key factors to improve the success rate of surgery and reduce postoperative complications. Especially in kidney surgery, accurate surgical path planning is particularly important because the kidney is deep in the body, the surrounding structure is complex and the function is important.

[0003] Traditional surgical path planning relies on the doctor's personal experience and preoperative imaging examinations such as CT or MRI scans.

[0004] However, the above methods have many limitations, including high dependence on physician experience, subjectivity in image interpretation, and lack of real-time adjustment capabilities. They are subjective and uncertain, so there is room for improvement. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and propose an artificial intelligence-based kidney surgery auxiliary path planning system. Its advantages are that it significantly improves the accuracy and safety of the surgery and reduces the risk of accidental injuries and complications during the surgery.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] The artificial intelligence-based kidney surgery auxiliary path planning system includes an intelligent image recognition and processing module, a personalized surgical path planning module, a real-time surgical navigation and adjustment module, an integrated medical database and knowledge base, a data analysis and machine learning module, and a user-friendly interactive interface.

[0008] The method for using the artificial intelligence-based kidney surgery auxiliary path planning system includes the following steps:

[0009] Step 1: Use the intelligent image recognition and processing module to obtain kidney tumor surgery video data, extract frames and annotate the video. The processing stages include denoising, standardization and enhancement, generate a data set for deep learning training, and use deep learning algorithms to automatically identify key anatomical structures such as kidneys and tumors, and extract their features to obtain AI models for kidney, tumor and instrument recognition, which are used to accurately identify and locate organs and instruments.

[0010] Step 2: The personalized surgical path planning module uses decision trees, support vector machines, and random forest machine learning algorithms to automatically generate personalized surgical paths based on the patient's pathological conditions and surgical needs; optimizes the surgical path through simulated annealing and genetic algorithm optimization techniques; combines clinical experience and statistical methods to predict risk points and avoid them in surgical planning;

[0011] Step 3: The real-time surgical navigation and adjustment module obtains real-time image data by connecting to the real-time imaging equipment in the operating room. It uses optical or electromagnetic tracking technology to monitor the position and posture of surgical instruments in real time, and matches the real-time collected data with the preset surgical path, calculates the deviation, and adjusts the surgical path in time to adapt to the actual surgical environment.

[0012] Step 4: Use integrated medical databases and knowledge bases to integrate medical data from different sources, including surgical cases, pathological information, and technical guidelines, to build a comprehensive medical database, use semantic network knowledge representation methods to organize and encode the information in the database, and develop a rule-based and case-based decision support system to provide doctors with personalized surgical recommendations and training materials;

[0013] Step 5: Data Analysis and Machine Learning Module In terms of data mining, association rule learning and cluster analysis data mining techniques are applied to discover potential patterns and associations from historical surgical data. Model training is then performed, using gradient boosting and deep learning methods to train the optimization model for surgical path planning and navigation. The accuracy and generalization ability of the model are evaluated through cross-validation and A / B testing methods.

[0014] Step 6: User-friendly interactive interface adopts human-computer interaction design principles to create intuitive and easy-to-use interface layouts and elements; uses surgical path animation demonstration visualization tools to help doctors understand surgical planning and navigation information more intuitively; designs a real-time feedback system to collect doctors' feedback and suggestions, and continuously optimizes the interface design and interaction logic.

[0015] The present invention is further configured that the Gaussian filter used by the intelligent image recognition and processing module in the preprocessing stage is:

[0016]

[0017] Where G(x,y) is the value of the Gaussian filter at point (x,y) and σ is the standard deviation.

[0018] The present invention is further configured that the intelligent image recognition and processing module adopts a convolutional layer forward propagation form, and the propagation formula is:

[0019] Output=σ(Weight*Input+Bias)

[0020] Among them, σ represents the ReLU activation function, * represents the convolution operation, and this formula is used many times in image recognition and feature extraction for convolution calculation.

[0021] The present invention is further configured that the personalized surgical path planning module supports a vector machine for surgical path planning, and the optimization formula of the support vector machine is:

[0022]

[0023] subject toys (i) (w T x (i) +b)≥1-ξ i ,ξ i ≥0

[0024] Among them, w and b are the parameters of the hyperplane, C is the regularization parameter, ξ i is a slack variable.

[0025] The present invention is further configured that the real-time surgical navigation and adjustment module needs to perform path matching and adjustment, and the least squares optimization method is used here to reduce the error, and the formula is:

[0026]

[0027] Where A is the coefficient matrix, x is the unknown vector, and b is the observation vector.

[0028] The present invention is further configured such that when the integrated medical database and the knowledge base learn association rules, the support and confidence calculation formulas are respectively:

[0029] Support(X→Y)=P(X∪Y)

[0030] Confidence(X→Y)=P(Y|X)

[0031] Among them, X and Y are item sets, and P represents probability.

[0032] The present invention is further configured such that when the data analysis and machine learning module trains the deep learning model, the gradient calculation formula of the back propagation algorithm is:

[0033]

[0034] Where E is the loss function, is the weight between neurons i and j in layer l, is the weighted input to neuron i in layer l, is the error term of neuron i in layer l, is the activation output of neuron j in layer l-1.

[0035] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned artificial intelligence-based kidney surgery assisted path planning method when executing the computer program.

[0036] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned artificial intelligence-based kidney surgery assisted path planning method.

[0037] The beneficial effects of the present invention are:

[0038] 1. Significantly improve surgical accuracy and safety: Existing technologies often rely on the doctor's experience and manual measurement in surgical path planning, resulting in certain errors and uncertainties. However, the present invention integrates a deep learning model to automatically and accurately identify key anatomical structures and lesion areas such as kidneys, blood vessels, and tumors, providing an accurate basis for surgical path planning. This precise recognition capability significantly improves the accuracy and safety of surgery and reduces the risk of accidental injuries and complications during surgery.

[0039] 2. Optimize the surgical path and improve surgical efficiency: Traditional surgical path planning methods may lack systematicity and optimization, resulting in inefficient surgical paths or potential risks. The present invention combines optimization algorithms to generate a variety of feasible surgical path plans, and finds the best path through algorithm screening and optimization. This process not only considers the safety of the surgical path, but also takes into account surgical efficiency and patient prognosis, thereby improving the overall efficiency of the operation.

[0040] 3. Early warning of potential risks and reduction of complication rates: There are many potential risks during surgery, such as vascular damage, nerve compression, etc. Existing technologies often make it difficult to fully predict and warn of these risks. However, the present invention can detect and warn of potential risk factors in advance through path planning and risk assessment modules, and provide doctors with timely risk warnings and preventive measures. This helps reduce the incidence of surgical complications, improve the safety of surgery and the quality of patient recovery.

[0041] 4. Personalized surgical plan. The specific conditions and surgical needs of each patient are different. The existing technology often cannot provide a completely personalized surgical plan. However, the present invention can be customized according to the specific conditions of the patient and the needs of the doctor, and generate a surgical path plan that meets the characteristics of the patient. This personalized customization capability helps to achieve precision medicine and personalized treatment, and improve treatment effects and patient satisfaction.

[0042] 5. Shorten the surgical preparation time and improve the doctor's work efficiency: Traditional surgical path planning requires doctors to spend a lot of time on manual measurement and calculation, which increases the surgical preparation time. However, the present invention can significantly shorten the surgical preparation time through automation and intelligent processing. Doctors can obtain surgical path planning results more quickly and formulate surgical plans accordingly, thereby improving the doctor's work efficiency.

[0043] 6. Promote the progress of medical technology and improve the level of medical services: If the present invention can be successfully applied, it will not only solve the practical problems in kidney surgery path planning, but also promote the further development and application of artificial intelligence technology in the medical field. It will help to improve the level of intelligent medical services and provide patients with more efficient, accurate and safe medical services. At the same time, the innovation and practicality of the present invention also provide new ideas and methods for the progress and development of medical technology.

[0044] In summary, compared with the prior art, the artificial intelligence-based kidney surgery auxiliary path planning system of the present invention has brought significant improvements in the accuracy of surgical planning, real-time navigation and adaptability, the utilization efficiency of medical data, the doctor's operating experience and training effect. These improvements not only improve the success rate and safety of surgery, reduce surgical risks and complications, but also improve medical quality and efficiency, and enhance doctors' decision support and operational convenience. With the continuous advancement of technology and the expansion of application scenarios, it will bring more innovations and breakthroughs to the medical field. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flow chart of intelligent image recognition and processing of the artificial intelligence-based kidney surgery auxiliary path planning system proposed by the present invention;

[0046] Figure 2 A personalized surgical path planning flow chart of the artificial intelligence-based kidney surgery auxiliary path planning system proposed by the present invention;

[0047] Figure 3 A real-time surgical navigation and adjustment flow chart of the artificial intelligence-based kidney surgery auxiliary path planning system proposed by the present invention;

[0048] Figure 4 A flow chart of the integrated medical database and knowledge base of the artificial intelligence-based kidney surgery auxiliary path planning system proposed by the present invention;

[0049] Figure 5 The data analysis and machine learning flow chart of the artificial intelligence-based kidney surgery auxiliary path planning system proposed by the present invention;

[0050] Figure 6This is a flow chart of the user-friendly interactive interface of the artificial intelligence-based kidney surgery assisted path planning system proposed in the present invention. DETAILED DESCRIPTION

[0051] The technical solution of this patent is further described in detail below in conjunction with specific implementation methods.

[0052] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0053] Reference Figure 1-6 , an artificial intelligence-based kidney surgery auxiliary path planning system, including intelligent image recognition and processing module, personalized surgical path planning module, real-time surgical navigation and adjustment module, integrated medical database and knowledge base, data analysis and machine learning module and user-friendly interactive interface.

[0054] The method for using the artificial intelligence-based kidney surgery auxiliary path planning system includes the following steps:

[0055] Step 1: Use the intelligent image recognition and processing module to obtain kidney tumor surgery video data, extract frames and annotate the video. The processing stage includes denoising, standardization and enhancement, and generates a data set for deep learning training. Use deep learning algorithms to automatically identify key anatomical structures such as kidneys and tumors, and extract their features to obtain AI models for kidney, tumor and instrument recognition, which are used to accurately identify and locate organs and instruments. In the preprocessing stage, the Gaussian filter used is as follows:

[0056]

[0057] Where G(x,y) is the value of the Gaussian filter at point (x,y) and σ is the standard deviation.

[0058] The convolutional layer is used for forward propagation, and the propagation formula is:

[0059] Output=σ(Weight*Input+Bias)

[0060] Among them, σ represents the ReLU activation function, * represents the convolution operation, and this formula is used many times in image recognition and feature extraction for convolution calculation.

[0061] Advanced image recognition technology is used to deeply analyze and process real surgical videos before kidney surgery. Through artificial intelligence algorithms, the system can automatically identify kidneys, tumors, surrounding blood vessels and other key anatomical structures. This process not only improves the efficiency of image analysis, but also reduces errors caused by human factors. The system can automatically eliminate irrelevant information, focus on key structures, and provide accurate basic data for subsequent surgical path planning.

[0062] Step 2: The personalized surgical path planning module uses decision trees, support vector machines, and random forest machine learning algorithms to automatically generate personalized surgical paths based on the patient's pathological conditions and surgical needs; optimizes the surgical path through simulated annealing and genetic algorithm optimization techniques; combines clinical experience and statistical methods to predict risk points and avoid them in surgical planning; supports vector machines for surgical path planning, and the optimization formula for support vector machines is:

[0063]

[0064] subject toys (i) (w T x (i) +b)≥1-ξ i ,ξ i ≥0

[0065] Among them, w and b are the parameters of the hyperplane, C is the regularization parameter, ξ i is a slack variable.

[0066] Based on the results of identification and positioning, the system uses machine learning algorithms to automatically generate personalized surgical path planning. This plan takes into account surgical safety, efficiency, and patient-specific pathological conditions to provide doctors with the best surgical plan. The system will recommend the best surgical incision and operation path based on the size, location, morphology, and relationship of the tumor with surrounding tissues. At the same time, it can also predict possible risk points and plan in advance how to avoid important blood vessels and organs, thereby reducing surgical risks.

[0067] Step 3: The real-time surgical navigation and adjustment module obtains real-time image data by connecting to the real-time imaging equipment in the operating room; uses optical or electromagnetic tracking technology to monitor the position and posture of surgical instruments in real time, and matches the real-time collected data with the preset surgical path, calculates the deviation, and adjusts the surgical path in time to adapt to the actual surgical environment; path matching and adjustment are required, and the least squares optimization method is used here to reduce the error. The formula is:

[0068]

[0069] Where A is the coefficient matrix, x is the unknown vector, and b is the observation vector.

[0070] During the operation, the system receives and analyzes real-time image data by connecting to the imaging equipment in the operating room (such as endoscopes or robotic surgery systems) in real time. Using deep learning algorithms, the system can track the position of surgical instruments in real time and compare it with the pre-planned surgical path to provide real-time navigation for doctors. If unexpected situations occur during the operation, such as slight changes in the position of the tumor or unexpected bleeding, the system can immediately adjust the surgical path and instruct the doctor on how to respond. This function greatly improves the flexibility of the operation and the doctor's ability to respond.

[0071] Step 4: Use the integrated medical database and knowledge base to integrate medical data from different sources, including surgical cases, pathological information, and technical guidelines, to build a comprehensive medical database, use semantic network knowledge representation methods to organize and encode the information in the database, and develop a rule-based and case-based decision support system to provide doctors with personalized surgical advice and training materials; when learning association rules, the support and confidence calculation formulas are:

[0072] Support(X→Y)=P(X∪Y)

[0073] Confidence(X→Y)=P(Y|X)

[0074] Among them, X and Y are item sets, and P represents probability.

[0075] It integrates a rich medical database and knowledge base, including a large number of surgical cases, pathological data and surgical technical guidelines. These data not only provide a reference for surgical planning, but also provide decision-making support for doctors. Doctors can check the surgical plans and pathological results of similar cases at any time, refer to the standardized surgical procedures and technical guidelines provided by the system, and combine their own experience and judgment to develop the most appropriate surgical plan. This helps to improve the success rate of surgery and reduce complications.

[0076] Step 5: Data Analysis and Machine Learning Module In terms of data mining, association rule learning and cluster analysis data mining techniques are applied to discover potential patterns and associations from historical surgical data. Model training is then performed, using gradient boosting and deep learning methods to train the optimization model for surgical path planning and navigation. The accuracy and generalization ability of the model are evaluated through cross-validation and A / B testing methods. When training the deep learning model, the gradient calculation formula of the back propagation algorithm is:

[0077]

[0078] Where E is the loss function, is the weight between neurons i and j in layer l, is the weighted input to neuron i in layer l, is the error term of neuron i in layer l, is the activation output of neuron j in layer l-1.

[0079] It has powerful data analysis and machine learning capabilities. It can collect and analyze the data of each operation, including surgical path, instrument movement trajectory, operation time, blood loss and other indicators to evaluate the surgical effect and doctor performance. Through deep learning algorithms, the system can continuously optimize its own surgical path planning and navigation functions to improve the accuracy and practicality of planning. At the same time, the system can also provide doctors with personalized surgical training and simulation training functions to help doctors improve their surgical skills and level.

[0080] Step 6: User-friendly interactive interface: Adopt the principles of human-computer interaction design to create intuitive and easy-to-use interface layout and elements; use surgical path animation demonstration visualization tools to help doctors understand surgical planning and navigation information more intuitively; design a real-time feedback system to collect doctors' feedback and suggestions, and continuously optimize the interface design and interaction logic;

[0081] It provides an intuitive, user-friendly interactive interface that allows doctors to easily interact with the system. Doctors can input specific requirements for surgical plans through the interface, view the surgical path generated by the system, and manually adjust the path when necessary. The interface design fully considers the doctor's operating habits and the needs of the surgical process, ensuring that doctors can quickly get started and use the system effectively. At the same time, the system also provides a wealth of visualization tools, such as surgical path animation demonstrations, to help doctors understand surgical planning and navigation information more intuitively.

[0082] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned artificial intelligence-based kidney surgery assisted path planning method when executing the computer program.

[0083] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned artificial intelligence-based kidney surgery assisted path planning method.

[0084] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An artificial intelligence-based kidney surgery auxiliary path planning system, characterized in that: It includes intelligent image recognition and processing module, personalized surgical path planning module, real-time surgical navigation and adjustment module, integrated medical database and knowledge base, data analysis and machine learning module and user-friendly interactive interface.

2. The method for using the artificial intelligence-based kidney surgery auxiliary path planning system according to claim 1, characterized in that: The following steps are involved: Step 1: Use the intelligent image recognition and processing module to obtain kidney tumor surgery video data, extract frames and annotate the video. The processing stages include denoising, standardization and enhancement, generate a data set for deep learning training, and use deep learning algorithms to automatically identify key anatomical structures such as kidneys and tumors, and extract their features to obtain AI models for kidney, tumor and instrument recognition, which are used to accurately identify and locate organs and instruments. Step 2: The personalized surgical path planning module uses decision trees, support vector machines, and random forest machine learning algorithms to automatically generate personalized surgical paths based on the patient's pathological conditions and surgical needs; optimizes the surgical path through simulated annealing and genetic algorithm optimization techniques; combines clinical experience and statistical methods to predict risk points and avoid them in surgical planning; Step 3: The real-time surgical navigation and adjustment module obtains real-time image data by connecting to the real-time imaging equipment in the operating room. It uses optical or electromagnetic tracking technology to monitor the position and posture of surgical instruments in real time, and matches the real-time collected data with the preset surgical path, calculates the deviation, and adjusts the surgical path in time to adapt to the actual surgical environment. Step 4: Use integrated medical databases and knowledge bases to integrate medical data from different sources, including surgical cases, pathological information, and technical guidelines, to build a comprehensive medical database, use semantic network knowledge representation methods to organize and encode the information in the database, and develop a rule-based and case-based decision support system to provide doctors with personalized surgical recommendations and training materials; Step 5: Data Analysis and Machine Learning Module In terms of data mining, association rule learning and cluster analysis data mining techniques are applied to discover potential patterns and associations from historical surgical data. Model training is then performed, using gradient boosting and deep learning methods to train the optimization model for surgical path planning and navigation. The accuracy and generalization ability of the model are evaluated through cross-validation and A / B testing methods. Step 6: User-friendly interactive interface adopts human-computer interaction design principles to create intuitive and easy-to-use interface layouts and elements; uses surgical path animation demonstration visualization tools to help doctors understand surgical planning and navigation information more intuitively; designs a real-time feedback system to collect doctors' feedback and suggestions, and continuously optimizes the interface design and interaction logic.

3. The method for using the artificial intelligence-based kidney surgery auxiliary path planning system according to claim 2, characterized in that: The Gaussian filter used by the intelligent image recognition and processing module in the preprocessing stage is as follows: Where G(x,y) is the value of the Gaussian filter at point (x,y) and σ is the standard deviation.

4. The method for using the artificial intelligence-based kidney surgery auxiliary path planning system according to claim 3 is characterized in that: The intelligent image recognition and processing module adopts the forward propagation form of the convolution layer, and the propagation formula is: Output=σ(Weight*Input+Bias) Among them, σ represents the ReLU activation function, * represents the convolution operation, and this formula is used many times in image recognition and feature extraction for convolution calculation.

5. The method for using the artificial intelligence-based kidney surgery auxiliary path planning system according to claim 4 is characterized in that: The personalized surgical path planning module supports a vector machine for surgical path planning, and the optimization formula of the support vector machine is: subject toy (i) (w T x (i) +b)≥1-ξ i ,ξ i ≥0 Among them, w and b are the parameters of the hyperplane, C is the regularization parameter, ξ i is a slack variable.

6. The method for using the artificial intelligence-based kidney surgery auxiliary path planning system according to claim 5, characterized in that: The real-time surgical navigation and adjustment module needs to perform path matching and adjustment. The least squares optimization method is used here to reduce the error. The formula is: Where A is the coefficient matrix, x is the unknown vector, and b is the observation vector.

7. The method for using the artificial intelligence-based kidney surgery auxiliary path planning system according to claim 6, characterized in that: When the integrated medical database and knowledge base learn association rules, the support and confidence calculation formulas are: Support(X→Y)=P(X∪Y) Confidence(X→Y)=P(Y|X) Among them, X and Y are item sets, and P represents probability.

8. The method for using the artificial intelligence-based kidney surgery auxiliary path planning system according to claim 7, characterized in that: When the data analysis and machine learning module trains the deep learning model, the gradient calculation formula of the back propagation algorithm is: Where E is the loss function, is the weight between neurons i and j in layer l, is the weighted input to neuron i in layer l, is the error term of neuron i in layer l, is the activation output of neuron j in layer l-1.

9. 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 processor executes the computer program, the steps of the method according to any one of claims 2 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 2 to 8 are implemented.

Images