An artificial intelligence-based puncture navigation system and method for egg retrieval robot

Abstract

The present invention belongs to the technical field of ovarian puncture, and specifically relates to an artificial intelligence-based egg retrieval robot puncture navigation system and method. The invention determines the distribution position of mature follicles according to the reference image, thereby selecting the sampling follicles that need to be punctured, and based on the sampling follicles, the puncture path planned in advance realizes accurate navigation of the puncture process of the egg retrieval robot. During the puncture process, the present invention collects the force feedback information of the puncture needle in real time, including resistance parameters and penetration parameters, and adjusts the pre-execution puncture force of the puncture needle in real time by comparing the current penetration parameter with the required penetration force to ensure the smoothness and accuracy of the puncture process. At the same time, the risk score of the puncture is determined according to the real-time puncture position and pre-execution puncture force of the puncture needle, so as to evaluate the executableness of the puncture path, and automatically switch to the backup puncture path when the puncture path is not executable, so as to ensure the safety of the egg retrieval puncture process.

Description

Technical Field

[0001] The present invention belongs to the technical field of ovarian puncture, and in particular relates to an artificial intelligence-based egg retrieval robot puncture navigation system and method. Background Art

[0002] With the continuous advancement of medical technology, more and more robot-assisted technologies are being used in clinical operations. In assisted reproductive technology, egg retrieval surgery is an important link. The success of this operation is directly related to the success rate of subsequent embryo culture and transplantation. Traditional egg retrieval surgery mainly relies on the doctor's experience and feel, but this method has problems such as high difficulty in operation, low accuracy, and high risk. Therefore, how to provide an egg retrieval robot puncture navigation system and method that can accurately navigate and reduce surgical risks has become a problem that needs to be solved urgently.

[0003] In the prior art, there are already some robot systems for assisting egg retrieval surgery, but they often only focus on minimizing damage to other tissues during the puncture navigation process. However, it is inevitable that some tissue structures will be penetrated by the puncture needle, and the puncture force will be affected. However, excessive puncture force may cause relatively irreversible damage to some unpunctured or punctured tissue structures, which obviously cannot meet actual needs. Based on this, the present solution provides an egg retrieval robot puncture navigation method that can adjust the puncture force in real time to ensure the safety of the puncture process. Summary of the invention

[0004] The purpose of the present invention is to provide an artificial intelligence-based egg retrieval robot puncture navigation system and method, which can adjust the puncture force in real time based on the stress conditions of the tissue structure during the puncture process, thereby ensuring the safety and accuracy of the puncture process.

[0005] The technical solution adopted by the present invention is as follows:

[0006] An artificial intelligence-based puncture navigation method for an egg retrieval robot, comprising:

[0007] Acquiring medical imaging data of the ovary, preprocessing the medical imaging data, and calibrating the preprocessed medical imaging data as a reference image;

[0008] Outputting the distribution positions of mature follicles in the ovaries according to the reference image, selecting sampling follicles that need to be punctured according to the distribution positions of the follicles, and recording multiple puncture points on the sampling follicles;

[0009] Delineating the puncture area, planning multiple puncture paths according to the puncture point and the puncture area, and outputting the preferred puncture path and the backup puncture path;

[0010] Puncturing the follicle based on the preferred puncture path, and collecting force feedback information of the puncture needle during the puncture process in real time;

[0011] The pre-execution puncture force of the puncture needle is output in real time according to the force feedback information, and the risk score of the puncture is determined based on the real-time puncture position of the puncture needle and the pre-execution puncture force, and the feasibility of the preferred puncture path is output according to the risk score, and when the preferred puncture path is not executable, the backup puncture path is switched to continue the puncture.

[0012] In a preferred embodiment, the steps of obtaining medical image data of the ovary, preprocessing the medical image data, and calibrating the preprocessed medical image data as a reference image include:

[0013] Acquire the ultrasonic image data of the ovary, and perform denoising processing to obtain a denoised image;

[0014] Delineating the ovarian region from the denoised image, and extracting an image of the ovarian region from the denoised image;

[0015] The image of the ovarian region is three-dimensionally reconstructed to generate a three-dimensional image of the ovary, and the three-dimensional image is calibrated as a reference image.

[0016] In a preferred embodiment, the step of outputting the distribution positions of mature follicles in the ovaries based on the reference image and selecting the sampling follicles to be punctured according to the distribution positions of the follicles comprises:

[0017] Acquire a reference image, construct a virtual coordinate system in the reference image, and locate the distribution coordinates of all follicles according to the virtual coordinate system;

[0018] According to the distribution coordinates, the number of adjacent follicles of each follicle is determined and recorded as a first condition parameter, and then the distance between each follicle and the adjacent follicle is calculated and marked as a second condition parameter;

[0019] An evaluation function is obtained, and the first condition parameter and the second condition parameter are input into the evaluation function together, and the output result of the evaluation function is calibrated as an evaluation score, and then the follicle with the largest evaluation score is determined as the sampling follicle.

[0020] In a preferred embodiment, the step of recording a plurality of puncture points on the sampled follicles comprises:

[0021] Obtaining the geometric center point of the sampled follicle, setting an initial puncture point at the geometric center point, and setting a plurality of circular trajectories with different radii with the geometric center point as the center of the circle;

[0022] Evenly distributing a plurality of candidate puncture points on each of the annular trajectories, and calculating the shortest distance between each candidate puncture point and the edge of the sampled follicle, and marking it as a safe distance;

[0023] Obtaining a safety distance threshold, and comparing the safety distance with the safety distance threshold;

[0024] If the safety distance is less than the safety distance threshold, deleting the candidate puncture point corresponding to the safety distance;

[0025] If the safety distance is greater than or equal to the safety distance threshold, the candidate puncture point corresponding to the safety distance is recorded as the puncture point.

[0026] In a preferred embodiment, the steps of defining the puncture area, planning multiple puncture paths according to the puncture point and the puncture area, and outputting the preferred puncture path and the backup puncture path include:

[0027] Acquiring an image of the tissue structure between the punctured area and the punctured follicle;

[0028] Obtaining a standard puncture angle, and extending a virtual path within the standard puncture angle toward the puncture area starting from the puncture point;

[0029] Obtaining the number of tissue structures penetrated by the puncture needle under each of the virtual paths, and extracting virtual paths whose penetration numbers are less than or equal to the allowed penetration number as allowed puncture paths according to a preset allowed penetration number;

[0030] Obtaining the shortest distance between the puncture needle and the adjacent tissue structure under each of the allowable puncture paths, and recording it as a screening condition parameter;

[0031] The screening condition parameters are arranged according to their values, and the permissible puncture path with the largest value of the screening condition parameter is marked as the preferred puncture path, and the other permissible puncture paths are marked as alternative puncture paths.

[0032] In a preferred embodiment, the force feedback information includes resistance parameters and penetration force parameters of the puncture needle;

[0033] Wherein, the penetration parameter is always greater than the resistance parameter.

[0034] In a preferred embodiment, the step of outputting the pre-execution puncture force of the puncture needle in real time according to the force feedback information comprises:

[0035] Obtain the current penetration force parameters of the puncture needle in real time;

[0036] Obtaining the required penetration force of the tissue that has not been penetrated in the preferred puncture path;

[0037] comparing the current penetration parameter with the required penetration;

[0038] If the current penetration force parameter is greater than or equal to the required penetration force, continue to perform the puncture operation with the current penetration force parameter;

[0039] If the current penetration parameter is less than the required penetration, the difference between the current penetration parameter and the required penetration is calculated, and the current penetration parameter is increased according to the difference, and the required penetration is recorded as the pre-execution puncture force;

[0040] After penetrating the tissue that has not been penetrated, the friction force between the tissue and the puncture needle is output as a new resistance parameter, and the penetration force parameter of the puncture needle is adjusted again according to the new resistance parameter.

[0041] In a preferred embodiment, the step of determining the risk score of the puncture based on the real-time puncture position of the puncture needle and the pre-execution puncture force, and outputting the feasibility of the preferred puncture path according to the risk score, comprises:

[0042] Obtaining the puncture depth of the puncture needle at the real-time puncture position and the pre-execution puncture force at the real-time puncture position;

[0043] In the same mapping space, mapping processing is performed on the puncture depth and the pre-execution puncture force to obtain a first mapping parameter corresponding to the puncture depth and a second mapping parameter corresponding to the pre-execution puncture force;

[0044] Obtaining a scoring function, inputting the first mapping parameter and the second mapping parameter into the scoring function, and marking the output result thereof as a risk score;

[0045] An evaluation threshold is obtained and compared with the risk score. When the risk score is greater than or equal to the evaluation threshold, the preferred puncture path is determined to be unexecutable, the puncture needle is withdrawn, and the backup puncture path is switched. Otherwise, the puncture is continued along the preferred puncture path.

[0046] The present invention also provides an artificial intelligence-based egg retrieval robot puncture navigation system, which is applied to the artificial intelligence-based egg retrieval robot puncture navigation method, comprising:

[0047] An image acquisition module, the image acquisition module is used to acquire medical image data of the ovary, pre-process the medical image data, and calibrate the pre-processed medical image data as a reference image;

[0048] A puncture point generation module, the puncture point generation module is used to output the distribution position of mature follicles in the ovary according to the reference image, select the sampling follicles that need to be punctured according to the distribution position of the follicles, and record multiple puncture points on the sampling follicles;

[0049] A path planning module, the path planning module is used to define the puncture area, and plan multiple puncture paths according to the puncture point and the puncture area, and output a preferred puncture path and a backup puncture path;

[0050] A force feedback module, the force feedback module is used to perform puncture on the follicle based on the preferred puncture path, and collect force feedback information of the puncture needle during the puncture process in real time;

[0051] A risk assessment module, wherein the risk assessment module is used to output the pre-execution puncture force of the puncture needle in real time according to the force feedback information, and determine the risk score of the puncture based on the real-time puncture position of the puncture needle and the pre-execution puncture force, and output the feasibility of the preferred puncture path according to the risk score, and when the preferred puncture path is not executable, switch to the backup puncture path to continue the puncture.

[0052] And, an electronic device, the electronic device comprising:

[0053] at least one processor;

[0054] and a memory communicatively coupled to the at least one processor;

[0055] Wherein, the memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the above-mentioned artificial intelligence-based egg retrieval robot puncture navigation method.

[0056] The technical effects achieved by the present invention are:

[0057] The present invention determines the distribution position of mature follicles according to the reference image, thereby selecting the sampling follicles that need to be punctured, and based on the sampled follicles, plans the puncture path in advance, thereby realizing accurate navigation of the puncture process of the egg retrieval robot. During the puncture process, the present invention collects the force feedback information of the puncture needle in real time, including resistance parameters and penetration parameters, and adjusts the pre-execution puncture force of the puncture needle in real time by comparing the current penetration parameter with the required penetration force, thereby ensuring the smoothness and accuracy of the puncture process. At the same time, the puncture risk score is determined according to the real-time puncture position and pre-execution puncture force of the puncture needle, thereby evaluating the executableness of the puncture path, and automatically switching to the backup puncture path when the puncture path is not executable, thereby ensuring the safety of the egg retrieval puncture process. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 is a flow chart of the method of the present invention;

[0059] Figure 2 It is a system module diagram of the present invention;

[0060] Figure 3 It is a structural diagram of an electronic device of the present invention. DETAILED DESCRIPTION

[0061] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0062] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0063] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure or characteristic that may be included in at least one implementation of the present invention. The phrase "in a preferred embodiment" that appears in different places in this specification does not refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0064] See also Figure 1 As shown, the present invention provides an artificial intelligence-based egg retrieval robot puncture navigation method, comprising:

[0065] S1, obtaining medical imaging data of the ovary, preprocessing the medical imaging data, and calibrating the preprocessed medical imaging data as a reference image;

[0066] S2. Output the distribution positions of mature follicles in the ovary according to the reference image, select the sampling follicles that need to be punctured according to the distribution positions of the follicles, and record multiple puncture points on the sampling follicles;

[0067] S3, delineating the puncture area, planning multiple puncture paths according to the puncture point and the puncture area, and outputting the preferred puncture path and the backup puncture path;

[0068] S4, puncturing the follicle based on the preferred puncture path, and collecting force feedback information of the puncture needle during the puncture process in real time;

[0069] S5. Output the pre-execution puncture force of the puncture needle in real time according to the force feedback information, determine the risk score of the puncture based on the real-time puncture position of the puncture needle and the pre-execution puncture force, output the feasibility of the preferred puncture path according to the risk score, and switch to the backup puncture path to continue the puncture when the preferred puncture path is not executable.

[0070] As described in the above steps S1-S5, egg retrieval is a key step in procedures such as in vitro fertilization and intracytoplasmic sperm injection. The traditional egg retrieval process usually requires doctors to obtain eggs through manual puncture under the guidance of vaginal ultrasound. This process not only requires extremely high experience and skills of doctors, but also is a huge physical and psychological burden on patients. In this embodiment, medical imaging data of the ovaries are first obtained to provide data support for subsequent puncture navigation. The acquired medical imaging data will be pre-processed. The pre-processing process mainly includes operations such as denoising and contrast enhancement, the purpose of which is to improve the quality of the image. Then, the pre-processed medical imaging data is calibrated as a reference image to facilitate subsequent identification and positioning of follicles. On this basis, the distribution position of mature follicles in the ovaries is output according to the reference image. After the distribution position of the follicles is obtained, the sampling follicles that need to be punctured can be selected according to the distribution. After the sampling follicles are selected, multiple puncture points need to be recorded on the follicles to facilitate the subsequent planning of the puncture path. Then, the puncture area is delineated, and multiple puncture paths are planned based on the puncture points and the puncture area to ensure that the puncture is performed. The safety and accuracy of the puncture process avoids unnecessary damage to the follicles and surrounding tissues. The planning of the puncture path usually takes into account factors such as the coordinates of the puncture point and the size of the puncture tool. Based on the planned puncture path, the puncture operation can be performed. During the puncture process, the force feedback information of the puncture needle will be collected in real time. The force feedback information includes the resistance parameter and the penetration parameter of the puncture needle. The penetration parameter is always greater than the resistance parameter. The safety of the puncture is evaluated on this basis. Specifically, according to the force feedback information of the puncture needle, the pre-execution puncture force of the puncture needle is output in real time, and the risk score of the puncture is determined according to the real-time puncture position and the pre-execution puncture force of the puncture needle. The higher the risk score, the greater the difficulty and risk of the puncture, and other tissues may be damaged during the puncture process. Based on this, this embodiment outputs the executability of the preferred puncture path according to the risk score. If the risk score of the preferred puncture path is too high, it will be determined that it is not executable. At this time, it is necessary to switch to the backup puncture path for puncture and egg retrieval, so as to ensure the smooth progress of the puncture, thereby reducing the risk of puncture and improving the success rate of puncture.

[0071] In a preferred embodiment, the steps of acquiring medical image data of the ovary, preprocessing the medical image data, and calibrating the preprocessed medical image data as a reference image include:

[0072] S101, obtaining ultrasound image data of the ovary, and performing denoising processing to obtain a denoised image;

[0073] S102, delineating the ovarian region from the denoised image, and extracting an image of the ovarian region from the denoised image;

[0074] S103, performing three-dimensional reconstruction on the image of the ovarian region to generate a three-dimensional image of the ovary, and calibrating the three-dimensional image as a reference image.

[0075] As described in the above steps S101-S103, after the medical imaging data of the ovaries are collected, they need to be denoised. The purpose of the denoising is to eliminate irrelevant noise in the image, improve the clarity of the image, and provide accurate data support for subsequent image processing. After the denoising, the ovarian region needs to be delineated from the denoised image, which can be achieved through image segmentation. After the ovarian region is delineated, the image of the ovarian region needs to be extracted from the denoised image. After the image extraction of the ovarian region is completed, the image of the ovarian region is three-dimensionally reconstructed to provide more intuitive ovarian structure information. Finally, the generated three-dimensional image of the ovary is calibrated as the reference image.

[0076] In a preferred embodiment, the steps of outputting the distribution positions of mature follicles in the ovaries according to the reference image and selecting the sampling follicles to be punctured according to the distribution positions of the follicles include:

[0077] S201, acquiring a reference image, constructing a virtual coordinate system in the reference image, and locating the distribution coordinates of all follicles according to the virtual coordinate system;

[0078] S202, determining the number of adjacent follicles of each follicle according to the distribution coordinates, and recording it as a first condition parameter, and then calculating the distance between each follicle and the adjacent follicle, and marking it as a second condition parameter;

[0079] S203, obtaining an evaluation function, and inputting the first condition parameter and the second condition parameter into the evaluation function together, and marking the output result of the evaluation function as an evaluation score, and then determining the follicle with the largest evaluation score as the sampling follicle.

[0080] As described in the above steps S201-S203, after the reference image is output, a virtual coordinate system is constructed on the reference image to more accurately locate the position of the follicle, and the distribution coordinates of all follicles in the image are analyzed based on the virtual coordinate system. According to the distribution coordinates of the follicles, the adjacent relationship between each follicle and other follicles can be determined. In this embodiment, when determining the adjacent relationship between follicles, a standard distance is preset. Only when the distance between adjacent follicles is less than or equal to the standard distance, they will be determined as adjacent follicles. By analogy, the number of adjacent follicles of each follicle can be determined, and the number of adjacent follicles is used as the first condition parameter, and the distance between each follicle and the adjacent follicle is used as the second condition parameter. Then, the evaluation function is called, and the first condition parameter (the number of adjacent follicles) and the second condition parameter (the distance between follicles) are input into the evaluation function together, wherein the expression of the evaluation function is: R=af 1 +bf 2 , where R represents the evaluation score, a and b represent the weight factors of the first condition parameter and the second condition parameter respectively, and f 1 and f 2 They represent the first condition parameter and the second condition parameter respectively. After the evaluation score is output, the follicle with the largest evaluation score will be directly determined as the sampling follicle.

[0081] In a preferred embodiment, the step of recording multiple puncture points on the sampled follicles comprises:

[0082] S204, obtaining the geometric center point of the sampled follicle, setting an initial puncture point on the geometric center point, and setting a plurality of circular trajectories with different radii with the geometric center point as the center of the circle;

[0083] S205, evenly distributing multiple candidate puncture points on each circular trajectory, calculating the shortest distance between each candidate puncture point and the edge of the sampled follicle, and marking it as a safe distance;

[0084] S206, obtaining a safety distance threshold, and comparing the safety distance with the safety distance threshold;

[0085] If the safety distance is less than the safety distance threshold, the candidate puncture point corresponding to the safety distance is deleted;

[0086] If the safety distance is greater than or equal to the safety distance threshold, the candidate puncture point corresponding to the safety distance is recorded as the puncture point.

[0087] As described in the above steps S204-S206, after the sampling follicle is determined, first, the geometric center point of the sampling follicle is obtained. Specifically, the smallest circumscribed cuboid of the sampling follicle is determined, and the center point of the cuboid is set as the geometric center point of the sampling follicle, and the position of the geometric center point is set as the initial puncture point. Then, with the geometric center point as the center of the circle, multiple circular trajectories with different radii are set, and multiple candidate puncture points are evenly distributed on each circular trajectory. Then, the shortest distance between these candidate puncture points and the edge of the sampling follicle is calculated, and it is marked as a safe distance to ensure that the puncture process will not cause damage to the follicle. In this process, a safety distance threshold is also required. The safety distance threshold is determined based on the patient's specific situation and the characteristics of the follicles. Finally, the safety distance is compared with the safety distance threshold. If the safety distance is less than the safety distance threshold, it means that the puncture point is too close to the edge of the follicle and may cause damage to the follicle. In this case, the candidate puncture point corresponding to the safety distance will be deleted to ensure the safety of the puncture. Conversely, if the safety distance is greater than or equal to the safety distance threshold, it means that the candidate puncture point is safe. In this case, the candidate puncture point corresponding to the safety distance will be recorded as the puncture point.

[0088] In a preferred embodiment, the steps of defining the puncture area, planning multiple puncture paths according to the puncture point and the puncture area, and outputting the preferred puncture path and the backup puncture path include:

[0089] S301, acquiring an image of the tissue structure between the punctured area and the punctured follicle;

[0090] S302, obtaining a standard puncture angle, and extending a virtual path within the standard puncture angle toward the puncture area starting from the puncture point;

[0091] S303, obtaining the number of tissue structures penetrated by the puncture needle under each virtual path, and extracting virtual paths whose penetration numbers are less than or equal to the allowed penetration number as allowed puncture paths according to the preset allowed penetration number;

[0092] S304, obtaining the shortest distance between the puncture needle and the adjacent tissue structure under each permissible puncture path, and recording it as a screening condition parameter;

[0093] S305 , arranging the screening condition parameters according to their values, and marking the permissible puncture path with the largest screening condition parameter value as the preferred puncture path, and marking the other permissible puncture paths as alternative puncture paths.

[0094] As described in the above steps S301-S305, the delineation of the extracorporeal puncture area can be determined based on the doctor's experience or historical operation records, and then the tissue structure image between the puncture area and the punctured follicle is obtained to ensure that the patient's specific situation can be clearly understood when planning the puncture path. Then, the standard puncture angle is obtained, and the puncture point is used as the starting point to extend a virtual path within the standard puncture angle to the puncture area to reduce the occurrence of puncture paths that do not meet the puncture requirements. After the virtual path is planned, the number of tissue structures penetrated by the puncture needle under each virtual path needs to be obtained, according to The preset allowable number of penetrations can extract virtual paths whose number of penetrations is less than or equal to the allowable number of penetrations, and these paths are regarded as allowable puncture paths. Then, the shortest distance between the puncture needle and the adjacent tissue structure under each allowable puncture path is obtained. This shortest distance can be used as a screening condition parameter to evaluate the quality of the puncture path. After all the screening condition parameters are obtained, they are arranged according to their values. On this basis, the allowable puncture path with the largest value of the screening condition parameter is marked as the preferred puncture path, and the remaining allowable puncture paths are regarded as alternative puncture paths.

[0095] In a preferred embodiment, the step of outputting the pre-execution puncture force of the puncture needle in real time according to the force feedback information includes:

[0096] S501, obtaining the current penetration force parameter of the puncture needle in real time;

[0097] S502, obtaining the required penetration force of the tissue that has not been penetrated in the preferred puncture path;

[0098] S503, comparing the current penetration parameter with the required penetration;

[0099] If the current penetration force parameter is greater than or equal to the required penetration force, the puncture operation continues to be performed with the current penetration force parameter;

[0100] If the current penetration parameter is less than the required penetration, the difference between the current penetration parameter and the required penetration is calculated, and the current penetration parameter is increased according to the difference, and the required penetration is recorded as the pre-execution puncture force;

[0101] After penetrating the tissue that has not been penetrated, the friction force between the tissue and the puncture needle is output as a new resistance parameter, and the penetration force parameter of the puncture needle is adjusted again according to the new resistance parameter.

[0102] As described in the above steps S501-S503, when performing the puncture operation, it is necessary to monitor the current penetration force parameter of the puncture needle in real time. During the puncture process, different tissues have different requirements for puncture force, so it is necessary to obtain the required penetration force of the tissue that has not been penetrated in the preferred puncture path. After obtaining the current penetration force parameter and the required penetration force, the two will be compared. If the current penetration force parameter is greater than or equal to the required penetration force, it means that the puncture process is normal, and the puncture operation can continue to be performed with the current penetration force. If the current penetration force parameter is less than the required penetration force, it is necessary to adjust the penetration force of the puncture needle. The specific operation is to calculate the difference between the current penetration force parameter and the required penetration force, and then adjust the penetration force according to the difference. The current penetration force parameter is increased by the value. At the same time, the required penetration force is recorded as the pre-execution puncture force to provide a reference for subsequent puncture operations. In the subsequent puncture process, after penetrating the tissue that has not been penetrated, the friction between the tissue and the puncture needle needs to be output as the newly added resistance parameter, and according to the newly added resistance parameter, the penetration force parameter of the puncture needle is adjusted again (that is, the penetration force before penetrating the tissue that has not been penetrated, plus the newly added resistance parameter as the subsequent execution penetration force). This is mainly because the resistance of tissue penetration is greater than its sliding resistance relative to the puncture needle. This adjustment of the penetration force can effectively control the speed of the puncture needle and reduce the puncture speed of the puncture needle, which correspondingly reduces the difficulty of puncture and makes the puncture process safer.

[0103] In a preferred embodiment, the step of determining the risk score of the puncture based on the real-time puncture position of the puncture needle and the pre-execution puncture force, and outputting the feasibility of the preferred puncture path according to the risk score includes:

[0104] S504, obtaining the puncture depth of the puncture needle at the real-time puncture position and the pre-execution puncture force at the real-time puncture position;

[0105] S505, mapping the puncture depth and the pre-puncture force in the same mapping space to obtain a first mapping parameter corresponding to the puncture depth and a second mapping parameter corresponding to the pre-puncture force;

[0106] S506, obtaining a scoring function, and inputting the first mapping parameter and the second mapping parameter into the scoring function, and marking the output result as a risk score;

[0107] S507, obtaining an evaluation threshold, and comparing the evaluation threshold with the risk score, and when the risk score is greater than or equal to the evaluation threshold, determining that the preferred puncture path is not executable, withdrawing the puncture needle, and switching to the backup puncture path, otherwise, continuing to perform the puncture along the preferred puncture path.

[0108] As described in the above steps S504-S507, after the pre-execution puncture force is output, the puncture depth of the puncture needle at the real-time puncture position is also obtained. These two parameters are the conditions for evaluating the puncture risk. In this embodiment, the puncture depth and the pre-execution puncture force are mapped in the same mapping space, so that a first mapping parameter corresponding to the puncture depth and a second mapping parameter corresponding to the pre-execution puncture force can be obtained, so as to provide quantitative data support for subsequent risk assessment, and then obtain a scoring function, and input the first mapping parameter and the second mapping parameter into the scoring function to calculate the risk score, wherein the expression of the scoring function is: U=k 1 Y 1 +k 2 Y 2 , where U represents the risk score, k 1 represents the weight coefficient of the first mapping parameter, Y 1 represents the first mapping parameter, k 2 represents the weight coefficient of the second mapping parameter, Y 2 Represents the second mapping parameter. After that, the evaluation threshold is obtained, and the risk score is compared with the evaluation threshold. If the risk score is greater than or equal to the evaluation threshold, it means that the risk of the puncture path is high. At this time, it should be determined that the puncture path is not executable. In this case, the puncture needle should exit the current puncture path and switch to the backup puncture path. On the contrary, if the risk score is less than the evaluation threshold, it means that the risk of the puncture path is within an acceptable range and the puncture operation can continue.

[0109] See also Figure 2 , an artificial intelligence-based egg retrieval robot puncture navigation system, applied to the artificial intelligence-based egg retrieval robot puncture navigation method, comprising:

[0110] An image acquisition module, the image acquisition module is used to acquire medical image data of the ovary, pre-process the medical image data, and calibrate the pre-processed medical image data as a reference image;

[0111] The puncture point generation module is used to output the distribution position of mature follicles in the ovary according to the reference image, select the sampling follicles that need to be punctured according to the distribution position of the follicles, and record multiple puncture points on the sampling follicles;

[0112] A path planning module is used to define the puncture area, plan multiple puncture paths according to the puncture point and the puncture area, and output the preferred puncture path and the backup puncture path;

[0113] A force feedback module, which is used to perform puncture on the follicle based on the preferred puncture path and collect force feedback information of the puncture needle during the puncture process in real time;

[0114] The risk assessment module is used to output the pre-execution puncture force of the puncture needle in real time based on the force feedback information, and determine the risk score of the puncture based on the real-time puncture position of the puncture needle and the pre-execution puncture force, and output the feasibility of the preferred puncture path according to the risk score, and when the preferred puncture path is not executable, switch to the backup puncture path to continue the puncture.

[0115] In the above, the system includes an image acquisition module, a puncture point generation module, a path planning module, a force feedback module and a risk assessment module. The image acquisition module is responsible for acquiring medical imaging data of the ovaries, such as ultrasound, CT, etc., and then preprocessing the medical imaging data, such as denoising, enhancement, etc., to improve the image quality and obtain a reference image. The puncture point generation module outputs the distribution position of mature follicles in the ovaries according to the reference image. On this basis, the sampling follicles that need to be punctured are selected, and multiple puncture points are recorded on the sampling follicles to provide data support for puncture path planning. The path planning module is responsible for delineating the puncture area and according to the puncture points and puncture Multiple puncture paths are planned in the region, and the preferred puncture path and the backup puncture path are output at the same time. The force feedback module performs the puncture operation on the follicle based on the preferred puncture path, and collects the force feedback information of the puncture needle during the puncture process in real time, providing data basis for subsequent risk assessment. The risk assessment module outputs the pre-execution puncture force of the puncture needle in real time according to the force feedback information, and determines the risk score of the puncture according to the real-time puncture position and pre-execution puncture force of the puncture needle. According to the risk score, the feasibility of the preferred puncture path is evaluated. When the preferred puncture path is not executable, the risk assessment module will automatically switch to the backup puncture path to ensure the smooth progress of the puncture operation.

[0116] See also Figure 3 , an electronic device, the electronic device comprising:

[0117] at least one processor;

[0118] and a memory communicatively coupled to the at least one processor;

[0119] Among them, the memory stores a computer program that can be executed by at least one processor, and the computer program is executed by at least one processor so that the at least one processor can execute the above-mentioned artificial intelligence-based egg retrieval robot puncture navigation method.

[0120] The electronic device also includes one or more input devices and one or more output devices. The input device is used to receive the user's operating instructions and input data, such as a keyboard, a mouse, a touch screen, etc. The output device is used to display the processing results and provide user feedback, such as a display, a printer, a speaker, etc. These devices communicate with the processor and the memory through appropriate interfaces, so that the user can interact with the electronic device to realize the operation and control of the puncture navigation method of the egg retrieval robot based on artificial intelligence.

[0121] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0122] The above is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications should also be considered as the protection scope of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention shall be implemented according to the conventional means in the art unless otherwise specified and limited.

Claims

1. An artificial intelligence-based puncture navigation system for egg retrieval robots, characterized in that: include: An image acquisition module, the image acquisition module is used to acquire medical image data of the ovary, pre-process the medical image data, and calibrate the pre-processed medical image data as a reference image; A puncture point generation module, the puncture point generation module is used to output the distribution position of mature follicles in the ovary according to the reference image, select the sampling follicles that need to be punctured according to the distribution position of the follicles, and record multiple puncture points on the sampling follicles; A path planning module, the path planning module is used to define the puncture area, and plan multiple puncture paths according to the puncture point and the puncture area, and output a preferred puncture path and a backup puncture path; A force feedback module, the force feedback module is used to perform puncture on the follicle based on the preferred puncture path, and collect force feedback information of the puncture needle during the puncture process in real time; a risk assessment module, the risk assessment module being used to output the pre-execution puncture force of the puncture needle in real time according to the force feedback information, and to determine the risk score of the puncture based on the real-time puncture position of the puncture needle and the pre-execution puncture force, and to output the feasibility of the preferred puncture path according to the risk score, and to switch to the backup puncture path to continue the puncture when the preferred puncture path is not executable; The step of obtaining medical image data of the ovary, preprocessing the medical image data, and calibrating the preprocessed medical image data as a reference image includes: Acquire the ultrasonic image data of the ovary, and perform denoising processing to obtain a denoised image; Delineating the ovarian region from the denoised image, and extracting an image of the ovarian region from the denoised image; Performing three-dimensional reconstruction on the image of the ovarian region to generate a three-dimensional image of the ovary, and calibrating the three-dimensional image as a reference image; Outputting the distribution position of mature follicles in the ovary according to the reference image, and selecting the sampling follicles that need to be punctured according to the distribution position of the follicles, includes: Acquire a reference image, construct a virtual coordinate system in the reference image, and locate the distribution coordinates of all follicles according to the virtual coordinate system; According to the distribution coordinates, the number of adjacent follicles of each follicle is determined and recorded as a first condition parameter, and then the distance between each follicle and the adjacent follicle is calculated and marked as a second condition parameter; Obtaining an evaluation function, and inputting the first condition parameter and the second condition parameter into the evaluation function together, and marking the output result of the evaluation function as an evaluation score, and then determining the follicle with the largest evaluation score as the sampling follicle; Said recording of multiple puncture points on said sampling follicles includes: Obtaining the geometric center point of the sampled follicle, setting an initial puncture point at the geometric center point, and setting a plurality of circular trajectories with different radii with the geometric center point as the center of the circle; Evenly distributing a plurality of candidate puncture points on each of the annular trajectories, and calculating the shortest distance between each candidate puncture point and the edge of the sampled follicle, and marking it as a safe distance; Obtaining a safety distance threshold, and comparing the safety distance with the safety distance threshold; If the safety distance is less than the safety distance threshold, deleting the candidate puncture point corresponding to the safety distance; If the safety distance is greater than or equal to the safety distance threshold, the candidate puncture point corresponding to the safety distance is recorded as the puncture point; The step of demarcating the puncture area, planning a plurality of puncture paths according to the puncture point and the puncture area, and outputting a preferred puncture path and a backup puncture path includes: Acquiring an image of the tissue structure between the punctured area and the punctured follicle; Obtaining a standard puncture angle, and extending a virtual path within the standard puncture angle toward the puncture area starting from the puncture point; Obtaining the number of tissue structures penetrated by the puncture needle under each of the virtual paths, and extracting virtual paths whose penetration numbers are less than or equal to the allowed penetration number as allowed puncture paths according to a preset allowed penetration number; Obtaining the shortest distance between the puncture needle and the adjacent tissue structure under each of the allowable puncture paths, and recording it as a screening condition parameter; Arrange the screening condition parameters according to their values, and mark the permissible puncture path with the largest value of the screening condition parameter as the preferred puncture path, and mark the other permissible puncture paths as alternative puncture paths; The force feedback information includes resistance parameters and penetration force parameters of the puncture needle; Wherein, the penetration parameter is always greater than the resistance parameter; The real-time output of the pre-execution puncture force of the puncture needle according to the force feedback information includes: Obtain the current penetration force parameters of the puncture needle in real time; Obtaining the required penetration force of the tissue that has not been penetrated in the preferred puncture path; comparing the current penetration parameter with the required penetration; If the current penetration force parameter is greater than or equal to the required penetration force, continue to perform the puncture operation with the current penetration force parameter; If the current penetration parameter is less than the required penetration, the difference between the current penetration parameter and the required penetration is calculated, and the current penetration parameter is increased according to the difference, and the required penetration is recorded as the pre-execution puncture force; After penetrating the tissue that has not been penetrated, the friction force between the tissue and the puncture needle is output as a new resistance parameter, and the penetration force parameter of the puncture needle is adjusted again according to the new resistance parameter; The step of determining a puncture risk score based on the real-time puncture position of the puncture needle and the pre-execution puncture force, and outputting the feasibility of the preferred puncture path according to the risk score, comprises: Obtaining the puncture depth of the puncture needle at the real-time puncture position and the pre-execution puncture force at the real-time puncture position; In the same mapping space, mapping processing is performed on the puncture depth and the pre-execution puncture force to obtain a first mapping parameter corresponding to the puncture depth and a second mapping parameter corresponding to the pre-execution puncture force; Obtaining a scoring function, inputting the first mapping parameter and the second mapping parameter into the scoring function, and marking the output result thereof as a risk score; An evaluation threshold is obtained and compared with the risk score. When the risk score is greater than or equal to the evaluation threshold, the preferred puncture path is determined to be unexecutable, the puncture needle is withdrawn, and the backup puncture path is switched. Otherwise, the puncture is continued along the preferred puncture path.

2. An electronic device, characterized in that: The electronic device comprises: at least one processor; a memory communicatively coupled to the at least one processor; And the artificial intelligence-based egg retrieval robot puncture navigation system as described in claim 1.

Images