CN-121999979-A - Knee joint operation simulation method and system based on AI technology

CN121999979ACN 121999979 ACN121999979 ACN 121999979ACN-121999979-A

Abstract

The invention discloses a knee joint operation simulation method and system based on AI technology, which relates to the operation simulation technical field, firstly, constructing a three-dimensional model through knee joint CT and MRI images of a patient, identifying each information of a lesion area, then acquiring operation type and operation scheme according to each information of the lesion area, then monitoring the distance between a doctor and the three-dimensional model in real time, analyzing and adjusting the position of the three-dimensional model, and secondly, predicting the operation steps of the doctor, generating each surgical instrument according to the prediction result, analyzing and adjusting each surgical instrument when generating each surgical instrument, and finally analyzing the surgical simulation process of the doctor, so that whether the doctor can proficiently perform the surgery or not can be known, and the authenticity of the surgical simulation and the precision of generating each surgical instrument are ensured.

Inventors

ZHAO LIANG
Liao Zheting
WANG ZIQI
CHEN YUFAN
WU DESHENG
Qiu Gengtao

Assignees

广州医科大学附属第一医院（广州呼吸中心）

Dates

Publication Date: 20260508
Application Date: 20260121

Claims (10)

1. The knee joint operation simulation method based on the AI technology is characterized by comprising the following steps: Step one, data acquisition, namely acquiring CT and MRI images of the knee joint of a patient from a database; Dividing the CT and MRI images of the knee joint of the patient by using an AI algorithm, extracting each tissue structure of the knee joint, generating a three-dimensional model of the knee joint, and identifying the lesion area and the area of the lesion area; step three, determining the operation type, namely acquiring an image of a knee joint lesion area of a patient from the CT and MRI images of the knee joint of the patient, analyzing the lesion type of the knee joint of the patient according to the image of the lesion area, and determining the operation type according to the lesion type; Step four, making a surgical scheme, namely acquiring the type of the operation, the position and the area of the lesion area, automatically generating the surgical scheme, and sending the surgical scheme to a doctor, wherein the doctor knows the surgical scheme within a preset time period; Step five, simulating an operation process, namely acquiring the distance between a doctor and a knee joint three-dimensional model in real time by using a sensor when the doctor performs operation simulation on the knee joint three-dimensional model of a patient, analyzing and adjusting the position of the knee joint three-dimensional model, simultaneously predicting the operation steps of the doctor, generating each operation instrument, and analyzing and adjusting each generated operation instrument when generating each operation instrument; step six, proficiency analysis, which is to collect operation data of doctors by using sensors in the simulation operation process of the doctors, analyze the simulation process of the doctors according to the operation data and judge whether the doctors are proficient in performing the operation.
2. The AI-technology-based knee surgery simulation method according to claim 1, wherein the determining of the surgery type is performed as follows: S11, acquiring an image of a knee joint lesion area, inputting the image into a trained learning model, outputting key features of the knee joint lesion area, comparing the key features of the knee joint lesion area with key features of knee joint lesion types in a database, and if the key features of the knee joint lesion area are the same as the key features of a certain knee joint lesion type in the database, representing that the lesion type of the knee joint lesion area of a patient is the knee joint lesion type in the database; S12, obtaining the lesion type of the lesion area of the knee joint of the patient, comparing the lesion type of the lesion area of the knee joint of the patient with the lesion types solved by the operation types of the knee joints in the database, and taking the operation type of the knee joint in the database as the operation type of the patient if the lesion type of the knee joint of the patient is the same as the lesion type solved by the operation type of a certain knee joint in the database.
3. The AI-technology-based knee surgery simulation method according to claim 1, wherein the automatically generating a surgery plan comprises the following specific procedures: s21, acquiring the operation type, the position and the area of a lesion area of a patient; S22, comparing the operation type of the patient with the operation types in all operation schemes in the database, and if the operation type in one operation scheme in the database is the same as the operation type of the patient, taking the operation scheme as an initial operation scheme, and acquiring all the initial operation schemes in the database by the method; s23, acquiring the position and the area of a lesion area in each initial operation scheme, and according to an analysis formula: Obtain the first Similarity coefficient of lesion area to lesion area of patient in initial surgical plan In the following Represents the first The location of the lesion in the initial surgical plan, Representing the location of the lesion area of the patient, Represents the first The area of the lesion in the initial surgical plan, Representing the area of the lesion area of the patient, 、 Weight coefficients representing the positions and the areas respectively, Represents a natural constant of the natural product, The numbers representing each of the initial surgical protocols, =1,2,3,.., , Representing the total number of initial surgical protocols, And Are all positive integers; S24, comparing similarity coefficients of the lesion area in each initial operation scheme and the lesion area of the patient, selecting the initial operation scheme with the maximum similarity coefficient, and taking the initial operation scheme as the operation scheme of the patient.
4. The AI-technology-based knee surgery simulation method according to claim 1, wherein the analyzing and adjusting the position of the knee three-dimensional model comprises the following steps: S31, when a doctor performs operation simulation on a three-dimensional model of the knee joint of a patient according to an operation scheme, acquiring the distance between the doctor and the three-dimensional model of the knee joint of the patient in real time by using a sensor; S32, acquiring a distance range between a doctor and a knee joint of a patient in an operating room from a database, and taking the distance range as a standard distance range; S33, obtaining the distance between the doctor and the knee joint three-dimensional model, comparing the distance with a standard distance unit, if the distance between the doctor and the knee joint is in a standard distance range, representing that the distance between the doctor and the knee joint three-dimensional model is proper, if the distance between the doctor and the knee joint three-dimensional model is not in a standard distance range, representing that the distance between the doctor and the knee joint three-dimensional model is unreasonable, and when the distance between the doctor and the knee joint three-dimensional model is unreasonable, adjusting the position of the knee joint three-dimensional model.
5. The AI-technology-based knee joint surgery simulation method according to claim 4, wherein the adjusting the position of the knee joint three-dimensional model comprises the following steps: S41, establishing a three-dimensional coordinate system by taking a doctor as an origin, acquiring the position of the doctor by using a sensor, and marking as ; S42, obtaining the maximum value and the minimum value of the standard distance range, taking a doctor as a center point, taking the minimum value of the standard distance range as the minimum radius, and taking the maximum value of the standard distance range as the maximum radius: , , In the middle of 、 Representing the minimum value of the standard distance range and the maximum value of the standard distance range respectively, A feasible solution representing the lower limit of the knee joint three-dimensional model coordinates, A feasible solution representing the upper limit of the coordinates of the three-dimensional model of the knee joint, and , , , , , ; S43, a feasible solution of the lower limit of the knee joint three-dimensional model coordinate and a feasible solution of the upper limit of the knee joint three-dimensional model coordinate are taken as a region surrounded by the three-dimensional coordinate system, the knee joint three-dimensional model is moved, the sensor is used for collecting the knee joint three-dimensional model coordinate, and when the knee joint three-dimensional model coordinate is in the coordinate range, the knee joint three-dimensional model is not moved any more.
6. The AI-technology-based knee joint surgery simulation method according to claim 1, wherein the operation steps of a doctor are predicted at the same time, and each surgical instrument is generated, and the specific procedures are as follows: S51, when a doctor performs operation simulation on a three-dimensional model of a knee joint of a patient according to an operation scheme, setting each acquisition time according to a preset time interval, acquiring an operation video of the doctor at each acquisition time by using a high-definition camera, and acquiring an operation step of the doctor at each acquisition time by using a convolutional neural network; S52, comparing the operation steps of the doctor at each acquisition time with each operation step in the operation scheme, if the operation step of the doctor at a certain acquisition time is the same as a certain operation step in the operation scheme, the operation step of the doctor at the acquisition time is the operation step in the operation scheme, and the operation step in the operation scheme is called as the current operation step; S53, acquiring the next operation step of the current operation step from the operation scheme, acquiring each operation instrument required by each next operation step from the operation scheme, and sequentially generating the operation instruments according to the use sequence of each operation instrument in the operation step in the operation scheme.
7. The AI-technology-based knee joint surgery simulation method according to claim 1, wherein the analyzing and adjusting the generated surgical instruments comprises the following steps: S61, acquiring all surgical instruments and the sizes of all the component parts in an operating room from a database, and taking the overall sizes and the sizes of all the component parts in the operating room as the standard overall sizes of all the surgical instruments and the standard sizes of all the component parts; s62, acquiring the built-in overall size of each surgical instrument and the built-in size of each component part from a database; s63, comparing the built-in overall size of each surgical instrument and the built-in size of each component with the standard overall size and the standard size of each component, and then: , In the middle of Represents the first The built-in overall size of the individual surgical instruments, Represents the first The standard overall dimensions of the individual surgical instruments, Represents the first In the surgical instrument The built-in dimensions of the individual component parts, Represents the first In the surgical instrument The standard dimensions of the individual component parts, Represents the first The accuracy of the individual surgical instruments returns a value, Representing the number of each surgical instrument, =1,2,3,.., , Representing the total number of surgical instruments, The numbers representing the respective constituent elements are given, =1,2,3,..., , Representing the total number of component parts, 、、 And Are all positive integers; S63, when the accuracy return value of a certain surgical instrument is1, representing that the generation of the surgical instrument meets the standard, and when the accuracy return value of the certain surgical instrument is 0, representing that the generation of the surgical instrument does not meet the standard, and judging whether the generation of each surgical instrument meets the standard or not by the method; S64, obtaining all surgical instruments which do not meet the standard, namely all the surgical instruments, comparing the built-in overall size of all the surgical instruments and the built-in size of all the component parts with the standard overall size of all the surgical instruments and the standard energy-saving size of all the component parts, if the built-in overall size of a certain surgical instrument is different, adjusting the built-in value of the overall size of the surgical instruments, and the adjusted built-in value is equal to the standard overall size value of the surgical instruments, and if the built-in size of a certain component part in a certain surgical instrument is different from the standard size of the component part in the surgical instruments, adjusting the built-in value of the component part in the surgical instruments, and the adjusted built-in value of the component part is equal to the standard size value of the component part, thereby adjusting the surgical instruments.
8. The AI-technology-based knee surgery simulation method according to claim 1, wherein the analyzing doctor is skilled in performing the surgery, and the specific procedures are as follows: When a doctor performs operation simulation on a three-dimensional model of a knee joint of a patient according to an operation scheme, a sensor is used for collecting operation data of the doctor, after the operation simulation is finished, operation data of the doctor in an operation simulation process are obtained, according to the operation data and standard data of operation operations in the operation scheme, simulation process coefficients of the doctor are analyzed, the simulation process coefficients of the doctor are compared with preset simulation process coefficient thresholds, if the simulation process coefficients of the doctor are lower than the preset simulation process coefficient thresholds, the doctor cannot be skillfully performed, and if the simulation process coefficients of the doctor are higher than the preset simulation process coefficients, the doctor can be skillfully performed.
9. The AI-technology-based knee surgery simulation method according to claim 8, wherein the analyzing doctor's simulation process coefficients are as follows: , In the middle of Represents the first The value of the data of the individual surgical operation, Represents the first The number of individual surgical procedure data in the surgical plan, A number representing the data of each surgical operation, =1,2,3,.., , Representing the total number of surgical operation data, Representing the simulated process coefficients of the physician, And Are all positive integers.
10. A knee surgery simulation system performing the AI-technology-based knee surgery simulation method according to any one of claims 1 to 9, comprising: the data acquisition module is used for acquiring CT and MRI images of the knee joint of the patient from the database; the three-dimensional reconstruction module is used for segmenting CT and MRI images of the knee joint of a patient by using an AI algorithm, extracting each tissue structure of the knee joint, generating a three-dimensional model of the knee joint by using a method, and identifying a lesion area; The operation type determining module is used for acquiring an image of a knee joint lesion area of a patient from the CT and MRI images of the knee joint of the patient, analyzing the lesion type of the knee joint of the patient according to the image of the lesion area, and determining the operation type according to the lesion type; The operation scheme making module is used for obtaining the type of the operation, the position and the area of the lesion area, automatically generating an operation scheme, and sending the operation scheme to a doctor, wherein the doctor knows the operation scheme within a preset time period; The operation process simulation module is used for acquiring the distance between a doctor and the knee joint three-dimensional model in real time by using a sensor when the doctor performs operation simulation on the knee joint three-dimensional model of a patient, analyzing and adjusting the position of the knee joint three-dimensional model, simultaneously predicting the operation steps of the doctor, generating each operation instrument, and analyzing and adjusting each generated operation instrument when each operation instrument is generated; The analysis module is used for acquiring operation data of a doctor by using a sensor in the simulation operation process of the doctor, analyzing the simulation process of the doctor according to the operation data, judging whether the doctor is skilled in performing the operation, and performing secondary operation simulation if the doctor is not skilled; the database is used to store various surgical protocols and various data in the operating room.

Description

Knee joint operation simulation method and system based on AI technology Technical Field The invention relates to the technical field of operation simulation, in particular to a knee joint operation simulation method and system based on an AI technology. Background The accurate medical treatment is a development trend of the modern medical industry, and before a patient performs knee joint operation, doctors repeatedly perform operation simulation operation, experience is accumulated in a virtual environment, and operation skills are improved, so that the operation is performed more gracefully and accurately in actual operation, operation risks are reduced, and operation success rate and patient satisfaction are improved. The traditional knee joint operation simulation method and system based on the AI technology have the following defects that 1, when a doctor performs an operation in an operating room, a certain distance exists between the doctor and the patient, and in the traditional knee joint operation simulation method and system based on the AI technology, the doctor does not monitor the distance between the doctor and the knee joint of the patient in real time and cannot accurately simulate the real operation environment when performing the operation simulation. 2. In an actual operating room, the sizes of all the surgical instruments are fixed, while in the traditional knee joint operation simulation method and system based on the AI technology, the sizes of all the surgical instruments are adjusted according to a three-dimensional model of the knee joint of a patient, and all the surgical instruments in the actual operating room cannot be accurately simulated. 3. In the conventional knee joint operation simulation method and system based on the AI technology, after the operation simulation of the doctor is finished, the simulation process of the doctor is not analyzed, and whether the doctor can perform the operation skillfully cannot be known. Disclosure of Invention In view of the above-mentioned technical shortcomings, the present invention aims to provide a knee joint operation simulation method and system based on AI technology. In order to solve the technical problems, the invention adopts the following technical scheme that the invention provides a knee joint operation simulation method based on an AI technology. And step two, three-dimensional reconstruction, namely segmenting the CT and MRI images of the knee joint of the patient by using an AI algorithm, extracting each tissue structure of the knee joint, generating a three-dimensional model of the knee joint, and identifying the lesion area and the area of the lesion area. Step three, determining the operation type, namely acquiring an image of a knee joint lesion area of a patient from the CT and MRI images of the knee joint of the patient, analyzing the lesion type of the knee joint of the patient according to the image of the lesion area, and determining the operation type according to the lesion type. And fourthly, making a surgical scheme, namely acquiring the type of the operation, the position and the area of the lesion area, automatically generating the surgical scheme, and sending the generated surgical scheme to a doctor, wherein the doctor knows the surgical scheme within a preset time period. And fifthly, simulating an operation process, namely acquiring the distance between a doctor and the knee joint three-dimensional model in real time by using a sensor when the doctor performs operation simulation on the knee joint three-dimensional model of the patient, analyzing and adjusting the position of the knee joint three-dimensional model, simultaneously predicting the operation steps of the doctor, generating each operation instrument, and analyzing and adjusting each generated operation instrument when each operation instrument is generated. Step six, proficiency analysis, which is to collect operation data of doctors by using sensors in the simulation operation process of the doctors, analyze the simulation process of the doctors according to the operation data and judge whether the doctors are proficient in performing the operation. S11, acquiring an image of a knee joint lesion area, inputting the image into a trained learning model, outputting key features of the knee joint lesion area, comparing the key features of the knee joint lesion area with key features of knee joint lesion types in a database, and if the key features of the knee joint lesion area are the same as the key features of a certain knee joint lesion type in the database, representing that the lesion type of the knee joint lesion area of a patient is the knee joint lesion type in the database. S12, obtaining the lesion type of the lesion area of the knee joint of the patient, comparing the lesion type of the lesion area of the knee joint of the patient with the lesion types solved by the operation types of the knee joints in the database, and taking the