Surgical skill simulation training method

By combining the skill training box, endoscopic lens and computer processing data, using image processing algorithms for intelligent evaluation, the problem of time-consuming and lack of tactile feedback in laparoscopic surgery training is solved, and the intelligence of surgical skills simulation training is realized and quantification, providing operational feedback similar to real surgery, helping users judge themselves and improve their skills.

CN118968839APending Publication Date: 2024-11-15SHANGHAI UNIV +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411205843.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-30
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

In the existing standardized training methods for laparoscopic surgery, the training results of the skill training box require manual evaluation by experts, which consumes manpower and time. The virtual reality training environment lacks the touch and feedback of surgical operations, and cannot provide a sense of reality close to the real surgery.

Method used

Using a combination method of skill training box, 1080p endoscopic lens and computer processing data, intelligent evaluation is carried out through the software interface real-time display of pictures and image processing algorithms, and quantitative evaluation of surgical skills is used by computer vision recognition algorithms to provide quantitative training results and self-select Judgment function.

Benefits of technology

实现了手术技能模拟训练的智能化、量化和数据化,减轻了专家评估负担,提供了与真实手术相近的操作反馈,方便用户自我评判和技能提升。

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118968839A_ABST
    Figure CN118968839A_ABST
Patent Text Reader

Abstract

The invention relates to a surgical skill simulation training method, which belongs to the field of surgery, and comprises the following three steps: S1, operating training of a skill training box, S2, capturing an operation picture by a 1080p endoscope lens, and S3, processing data by a computer. In the step S3 of processing the data by the computer, the computer is used as a carrier, and a picture S31 and an image processing algorithm quantitative evaluation S32 are displayed in real time through a software interface. According to the surgical skill simulation training method, on the basis of an existing skill training box, a software and hardware combination method is used, computer software and an algorithm are combined, the surgical skill simulation training method is achieved, the control ability of a user to operation pliers is trained, and in a step software interface, a picture S31 is displayed in real time, and the operation ability of the user is improved. A user observes an operation picture in real time through a software interface and adjusts the data of the motion of the operation clamp fed back through algorithm measurement, so that the user can judge the skill simulation training process of himself / herself conveniently.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of surgical technology, and in particular to a surgical skill simulation training method. Background Art

[0002] Laparoscopic surgery is a surgical treatment technology with distinctive characteristics and is considered a milestone in the history of surgical development. With the development of laparoscopic technology, many general surgical operations can now be completed through laparoscopic operations. Laparoscopic surgery for some diseases has even replaced traditional open surgery and become the gold standard surgical method. However, laparoscopic surgery has a long learning curve and is difficult, and the method of using patients as surgical technology training objects no longer meets the requirements of the development of the times. Standardized training of laparoscopic surgery is an important way to improve the level of laparoscopic surgical skills of physicians. Its purpose is to make surgical operations no longer rely solely on mechanical observation and imitation, but to form a standardized training system. At present, the laparoscopic surgery training systems on the market are mainly divided into virtual reality training boxes and surgical training boxes. Virtual reality training boxes can simulate complex surgical operations and are widely used. However, the traditional evaluation method of surgical skill simulation training is time-consuming, relies on personal judgment and is subjective. Using artificial intelligence for surgical evaluation is a more objective solution that reduces the burden on experts. For example, in 2022, Yilmaz proposed using deep neural networks in virtual reality simulations to continuously monitor surgical skill levels. The surgical training box targets specific training methods and projects, such as using surgical forceps to clamp and take operations to train hand-eye coordination. At the same time, Jin et al. proposed a surgical video tool detection and skill assessment method based on regional convolutional neural network.

[0003] The existing standardized training for laparoscopic surgery mainly adopts the training and evaluation of doctors' ability to manipulate surgical instruments, while practical operations allow doctors to perform surgical operations in simulated or real environments to improve their skills. However, the current standardized training methods for laparoscopic surgery still require experts to manually evaluate and score the training results in the skill training box, which is manpower and time-consuming. In addition, most virtual reality training environments lack the touch and feedback during the surgical operation and cannot provide a sense of reality close to the real operation. Therefore, there are shortcomings in the existing standardized training for laparoscopic surgery. This research topic proposes a surgical skill simulation training method to intelligently develop the ability of new doctors to manipulate surgical forceps, quantitatively identify lesions, and visually track surgical forceps during the standardized training process, combining quantitative training for daily operations with intelligent result evaluation functions that replace experts. Summary of the invention

[0004] 1. Technical issues to be resolved In view of the shortcomings of the existing technology, the present invention provides a surgical skill simulation training method with the advantages of intelligence, quantification, and dataization. It solves the problem that the training results of the skill training box need to be manually evaluated and scored by experts, which consumes manpower and time. In addition, most virtual reality training environments lack the touch and feedback during the surgical operation and cannot provide a sense of reality close to real surgery.

[0005] (II) Technical solution In order to achieve the above-mentioned quantification and dataization purposes, the present invention provides the following technical solutions: including three major steps: skill training box operation training S1, 1080p endoscope lens capture operation screen S2, and computer data processing S3; in the computer data processing S3, the computer is used as a carrier, and the real-time display screen S31 and image processing algorithm quantitative evaluation S32 are respectively composed through the software interface. The specific steps are as follows: Step 1 S1, the user logs in to a specific software system, grabs the target, and in the skill training box operation training S1, the user completes an operation clamp grabbing operation in the skill training box by holding the operation clamp to grab the target object on the 3D printing plate; Step 2 S2, the software system displays the operation screen in real time through the connected endoscope lens, the 1080p endoscope lens captures the operation screen S2, monitors the operation screen in the skill training box operation training S1 in real time, and transmits the screen to the computer in real time by connecting to the computer; Step three S3, intelligently analyze the surgical simulation training operation process and operation results through the designed computer vision recognition algorithm, and give quantitative results; in the computer processing data and training model S3, the input surgical image is accurately captured to the position information of the operating forceps through the trained algorithm model; the computer processing data S3 includes two steps: real-time display of the software interface S31 and quantitative evaluation of the image processing algorithm S32; The S31 software interface displays the screen in real time, and displays the skill simulation training images transmitted in the 1080p endoscope lens capture operation screen S2 in real time in the designed software interface. The computer inputs each frame of the training process into the improved artificial intelligence target detection algorithm, and quantitatively gives the skill training result data in the surgical skill simulation training; Preferably, the evaluation result is composed of four steps: calculating the clamping target size S321, calculating the clamping target distance S322, operating the clamp tracking S323, and quantitatively evaluating the skill training result S324. The specific steps are as follows: First, the clamping target size S321 and the clamping target distance S322 are calculated by using conventional image processing algorithms such as contour detection and reference objects of known size and distance; Second, step S324 of quantitatively evaluating the skill training results quantitatively calculates the skill training data through the operating forceps tracking results in S323: training completion time, operating forceps movement trajectory, maximum offset distance between the operating forceps and the target to be clamped, and other indicators, to achieve intelligent quantitative evaluation of the skill training results, and return the results to the software interface and display them to the user together with the skill training result image.

[0006] Preferably, in step three S3, both the clamping target and the reference object are geometric shapes made of 3D printing resin materials.

[0007] Preferably, in step three S3, in the operating forceps tracking S323, an improved YOLOv8 target detection neural network is obtained by adding a CFP module to the existing YOLOv8 target detection neural network, and the position information of the operating forceps is accurately captured from the input surgical image through the trained algorithm model.

[0008] (III) Beneficial effects Compared with the prior art, the present invention provides a surgical skill simulation training method, which has the following beneficial effects: 1. This surgical skill simulation training method uses a combination of software and hardware. On the basis of the existing skill training box, it combines computer software and algorithms to realize a surgical skill simulation training method to train users' ability to manipulate the operating forceps. In the step S31, the software interface displays the screen in real time. The user observes the operation screen in real time through the software interface, and makes adjustments based on the data of the movement of the operating forceps measured and fed back by the algorithm. At the same time, after successfully clamping the target, the screen will feedback the operation data of the skill simulation training process, the movement trajectory of the operating forceps, the training completion time, and the farthest distance between the operating forceps and the target to be clamped. It is convenient for users to self-evaluate their own skill simulation training process. This is a function lacking in traditional surgical skill simulation training methods.

[0009] 2. The surgical skill simulation training method, in step S32 of quantitative evaluation of the image processing algorithm, inputs each frame of the training process into the improved artificial intelligence target detection algorithm, quantitatively gives the skill training result data in the surgical skill simulation training, and converts the training process into visual quantitative data, which are surgical feature data with high correlation with the surgical skill level confirmed by existing research. This replaces the manual evaluation of the doctor's surgical skill simulation training level by experts. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A schematic diagram of the execution flow of a surgical skill simulation training method proposed by the present invention; Figure 2This is a schematic diagram of the quantitative evaluation process steps of the image processing algorithm for the surgical skill simulation training method proposed by the present invention; Figure 3 A schematic diagram of calculating the clamping target size according to a surgical skill simulation training method proposed by the present invention; Figure 4 A schematic diagram of calculating the clamping target distance according to a surgical skill simulation training method proposed by the present invention; Figure 5 This is a schematic diagram of the result output of a surgical skill simulation training method proposed in the present invention. DETAILED DESCRIPTION

[0011] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0012] See also Figure 1-5 , Embodiment 1: A surgical skill simulation training method, including skill training box operation training S1, 1080p endoscope lens capture operation screen S2, computer data processing S3 three steps; in the computer data processing S3, the computer is used as a carrier, the software interface real-time display screen S31 and the image processing algorithm quantitative evaluation S32 are respectively composed, and the specific steps are as follows: Step 1 S1, the user logs in to a specific software system, clamps the target size, and in the skill training box operation training S1, the user completes an operation clamp clamping operation in the skill training box by holding the operation clamp to clamp the target object on the 3D printing plate.

[0013] It should be noted that the model of the skill training box and the camera can be selected according to the required cost, etc. This step realizes the surgical skill simulation training without the need for virtual reality, reduces the technical difficulty, and reduces the cost.

[0014] Step two S2, the software system displays the operation screen in real time through the connected endoscopic lens, the 1080p endoscopic lens captures the operation screen S2, monitors the operation screen in the skill training box operation training S1 in real time, and transmits the screen to the computer in real time by connecting to the computer.

[0015] Step three S3, intelligently analyze the surgical simulation training operation process and operation results through the designed computer vision recognition algorithm, and give quantitative results; the computer processing data includes two steps: real-time display of the software interface S31 and quantitative evaluation of the image processing algorithm S32.

[0016] Among them, the real-time display screen S31 of the software interface displays the skill simulation training image transmitted in the 1080p endoscope lens capture operation screen S2 in real time in the designed software interface, so as to facilitate the subsequent display together with the evaluation results obtained in the image processing algorithm quantitative evaluation S32 in one interface.

[0017] Among them, quantitative evaluation of image processing algorithm S32 is the key step of this method, which consists of four steps: calculating the size of the clamping target S321, calculating the distance of the clamping target S322, tracking the clamp operation S323, and quantitatively evaluating the skill training results S324. First, the size of the clamping target is calculated S3219 (see Figure 3 ) and calculate the gripping target distance S322 (see Figure 4 ).

[0018] The step of quantitatively evaluating the skill training results S324 is to quantitatively calculate the skill training data through the operating forceps tracking results in the operating forceps tracking S323: training completion time, operating forceps movement trajectory, maximum offset distance between the operating forceps and the target to be clamped and other indicators, so as to realize intelligent quantitative evaluation of the skill training results, and return the results to the software interface for display together with the skill training result image.

[0019] It should be noted that both the clamping target and the reference object are geometric shapes made of 3D printing resin materials.

[0020] Among them, in the operating forceps tracking S323, the operating forceps are tracked using the improved YOLOv8 target detection neural network. In order to improve the accuracy of the operating forceps detection and reduce the interference and error caused by the overlap of the target and the operating forceps when clamping the target, the key information of the input data is focused on, thereby improving the performance of the model.

[0021] Among them, by adding the CFP module to the existing YOLOv8 target detection neural network ( Figure 5 ), to obtain the improved YOLOv8 target detection neural network, and accurately capture the position information of the operating forceps from the input surgical image through the trained algorithm model.

[0022] The beneficial effects of the present invention are: by combining with computer software and algorithms, a surgical skill simulation training method is realized to train the user's ability to manipulate the operating forceps. In the step of real-time display screen S31 of the software interface, the user observes the operation screen in real time through the software interface, and adjusts the data of the movement of the operating forceps fed back by the algorithm measurement. At the same time, after successfully clamping the target, the operation data of the skill simulation training process, the movement trajectory of the operating forceps, the training completion time, and the farthest distance between the operating forceps and the target to be clamped will be fed back on the screen. It is convenient for users to self-evaluate their own skill simulation training process. The computer inputs each frame image of the training process into the improved artificial intelligence target detection algorithm, quantitatively gives the skill training result data in the surgical skill simulation training, and converts the training process into visualized quantitative data. These data are surgical feature data that have been confirmed by existing studies to have a high correlation with the surgical skill level. This replaces the manual evaluation of the doctor's surgical skill simulation training level by experts.

[0023] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A surgical skill simulation training method, characterized in that: It includes three steps: skill training box operation training S1, 1080p endoscope lens capture operation screen S2, and computer data processing S3. In computer data processing S3, the computer is used as a carrier to display the screen in real time through the software interface S31 and the image processing algorithm quantitative evaluation S32. The specific steps are as follows: Step 1 S1, the user logs in to a specific software system, grabs the target, and in the skill training box operation training S1, the user completes an operation clamp grabbing operation in the skill training box by holding the operation clamp to grab the target object on the 3D printing plate; Step 2 S2, the software system displays the operation screen in real time through the connected endoscope lens, the 1080p endoscope lens captures the operation screen S2, monitors the operation screen in the skill training box operation training S1 in real time, and transmits the screen to the computer in real time by connecting to the computer; Step three S3, intelligently analyze the surgical simulation training operation process and operation results through the designed computer vision recognition algorithm, and give quantitative results; in the computer processing data and training model S3, the input surgical image is accurately captured to the position information of the operating forceps through the trained algorithm model; the computer processing data S3 includes two steps: real-time display of the software interface S31 and quantitative evaluation of the image processing algorithm S32; The S31 software interface displays the screen in real time, and the skill simulation training images transmitted in S2 captured by the 1080p endoscope lens are displayed in real time in the designed software interface. The computer inputs each frame of the training process into the improved artificial intelligence target detection algorithm, and quantitatively gives the skill training result data in the surgical skill simulation training.

2. A surgical skill simulation training method according to claim 1, characterized in that: The evaluation result consists of four steps: calculating the clamping target size S321, calculating the clamping target distance S322, operating the clamp tracking S323, and quantitatively evaluating the skill training result S324. The specific steps are as follows: First, the clamping target size S321 and the clamping target distance S322 are calculated by using conventional image processing algorithms such as contour detection and reference objects of known size and distance; Second, step S324 of quantitatively evaluating the skill training results quantitatively calculates the skill training data through the operating forceps tracking results in S323: training completion time, operating forceps movement trajectory, maximum offset distance between the operating forceps and the target to be clamped, and other indicators, to achieve intelligent quantitative evaluation of the skill training results, and return the results to the software interface and display them to the user together with the skill training result image.

3. A surgical skill simulation training method according to claim 1, characterized in that: Step 3 S3: Both the clamping target and the reference object are geometric shapes made of 3D printing resin materials.

4. A surgical skill simulation training method according to claim 1, characterized in that: In step three S3, in the operating forceps tracking S323, the CFP module is added to the existing YOLOv8 target detection neural network to obtain an improved YOLOv8 target detection neural network, and the position information of the operating forceps is accurately captured from the input surgical image through the trained algorithm model.