Bipolar electric coagulation forceps management method based on multi-data fusion and bipolar electric coagulation forceps

Abstract

The invention discloses a bipolar electric coagulation forceps management method based on multi-data fusion and bipolar electric coagulation forceps, and relates to the technical field of medical instruments. The invention discloses a bipolar electric coagulation forceps management method based on multi-data fusion. The method comprises the steps of bipolar electric coagulation forceps real-time data monitoring, bipolar electric coagulation forceps real-time management and abnormity monitoring and early warning. Patient physiological parameters, target affected area characteristics and image data are monitored in real time, and are subjected to multi-source data fusion analysis with bipolar coagulation forceps parameters, working environment and forceps tip acquisition signals, so that accurate extraction of real-time feedback characteristics and efficient detection of abnormal signals are realized; a management strategy is intelligently generated in combination with a real-time feedback management model, the bipolar coagulation forceps are dynamically optimized and adjusted, meanwhile, real-time early warning is provided based on an abnormal signal processing model, the safety and stability of equipment operation are guaranteed, and the intelligent level, the working efficiency and the clinical adaptability of the bipolar coagulation forceps are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical devices, and in particular to a bipolar electrocoagulation forceps management method based on multi-data fusion and a bipolar electrocoagulation forceps. Background Art

[0002] Traditional bipolar electrocoagulation forceps are an important tool in modern surgery. Their main function is to achieve precise and efficient vascular closure through electrocoagulation hemostasis. However, in complex surgeries such as cranial surgery, the functions of traditional electrocoagulation forceps are relatively single and cannot provide dynamic information of the target tissue, especially in cases involving changes in intracranial pressure. Patients lack real-time monitoring methods, which may increase the risk of brain injury and other postoperative complications. With the development of medical technology and microsensor technology, it has become possible to integrate multifunctional sensors into surgical instruments. High-precision pressure sensors, as an advanced monitoring tool, have been widely used in medical equipment to accurately measure pressure and provide real-time feedback. However, their application in bipolar electrocoagulation forceps is still in the exploratory stage.

[0003] Therefore, it is necessary to combine multi-data fusion technology to integrate data from multiple dimensions during the surgery to improve the management efficiency and accuracy of bipolar electrocoagulation forceps. Summary of the invention

[0004] The present invention aims to provide a bipolar electrocoagulation forceps management method and bipolar electrocoagulation forceps based on multi-data fusion, so as to improve the management efficiency and management accuracy of the bipolar electrocoagulation forceps.

[0005] A bipolar electrocoagulation forceps management method based on multi-data fusion comprises the following steps: Step S1: real-time data monitoring of bipolar electrocoagulation forceps; Acquire real-time monitoring data of patients; real-time monitoring data of patients includes physiological parameter data, characteristic data of target affected area and image data of affected area; At the same time, the real-time working data of the bipolar electrocoagulation forceps is obtained; the real-time working data of the bipolar electrocoagulation forceps includes the parameter data of the bipolar electrocoagulation forceps, the working environment data and the signal data collected by the forceps tip; According to the patient's real-time monitoring data and the real-time working data of the bipolar coagulation forceps, a real-time data analysis is performed using a bipolar coagulation forceps data fusion analysis model to obtain the real-time feedback characteristics of the bipolar coagulation forceps; wherein, when an abnormal feedback signal is identified in the real-time feedback characteristics of the bipolar coagulation forceps, step S3 is entered for abnormal processing; otherwise, the real-time feedback characteristics of the bipolar coagulation forceps are input into step S2 for analysis; the bipolar coagulation forceps data fusion analysis model realizes real-time and accurate analysis of multi-source data and intelligent feedback adjustment through multi-level data alignment, fusion and feature extraction, which effectively improves the working efficiency, adaptability and safety of the bipolar coagulation forceps; Step S2: real-time management of bipolar electrocoagulation forceps; The real-time feedback features of the bipolar coagulation forceps are input into the real-time feedback management model of the bipolar coagulation forceps for analysis to obtain the real-time management strategy of the bipolar coagulation forceps; the bipolar coagulation forceps are intelligently managed according to the real-time management strategy of the bipolar coagulation forceps; the real-time feedback management model of the bipolar coagulation forceps realizes the efficient linkage between the real-time feedback features and the management strategy through feature recognition, feature mapping based on the attention mechanism and the swarm optimization algorithm, and intelligent strategy output; Step S3: abnormal monitoring and early warning; When an abnormal feedback signal is received, an analysis is performed based on the real-time feedback characteristics of the bipolar coagulation forceps and the abnormal signal processing model of the bipolar coagulation forceps to obtain a real-time warning signal; subsequent measures are taken according to the real-time warning signal.

[0006] As a preferred technical solution of the present invention, the bipolar electrocoagulation forceps data fusion analysis model includes a data preprocessing layer, a multi-data fusion analysis layer, an abnormality judgment layer and a result output layer; The data preprocessing layer is used to time-align and preprocess the patient real-time monitoring data and the bipolar electrocoagulation forceps real-time working data using the DTW algorithm to obtain preprocessed patient real-time monitoring data and preprocessed bipolar electrocoagulation forceps real-time working data; The multi-data fusion analysis layer is used to perform data fusion analysis on the pre-processed patient real-time monitoring data and the pre-processed bipolar electrocoagulation forceps real-time working data to obtain the multi-data feedback characteristic value of the bipolar electrocoagulation forceps; The abnormal judgment layer is used to perform abnormal data analysis based on the multi-data feedback characteristic values ​​of the bipolar electrocoagulation forceps to obtain the abnormal data analysis result; if the abnormal data analysis result is abnormal, an abnormal feedback signal is output; otherwise, the output is empty; The result output layer is used to output the real-time feedback features of the bipolar coagulation forceps according to the multi-data feedback feature values ​​and abnormal data analysis results of the bipolar coagulation forceps.

[0007] As a preferred technical solution of the present invention, the specific steps of performing data fusion analysis in the multi-data fusion analysis layer include: In the multi-data fusion analysis layer, there are feature division layer, K convolution kernel layers H k and feature concatenation layer, k=1,2,…,K; The convolution kernel layer H k The convolution kernel in is divided into convolution kernel {H kn}, n=1, 2, ..., 4; convolution kernel {H kn}Equal dimensions between In the feature partitioning layer: Feature extraction is performed on the pre-processed patient real-time monitoring data and the pre-processed bipolar electrocoagulation forceps real-time working data to obtain the patient real-time monitoring data features and the bipolar electrocoagulation forceps real-time working data features; the patient real-time monitoring data features and the bipolar electrocoagulation forceps real-time working data features are spliced ​​and reorganized to obtain multi-data fusion features Z1 (k) and multi-data fusion features Z2 (k); in the feature division layer, k=0 at this time; In the convolution kernel layer H k middle: For the multi-data fusion feature Z1(k-1), the formula Z1(k)=Z1(k-1)*H is used. k1 Perform convolution operation to obtain multi-data fusion output feature Z1(k); Perform multi-resolution convolution transformation on the multi-data fusion feature Z2(k-1) to obtain the multi-data fusion output feature Z2(k); Specific steps: Using formula C k =avgPool(R k )[Z2(k-1)] performs pooling operation to obtain the pooled multi-data fusion output feature C k ; Among them, R k Represents the pooling step size; avgPool[] represents the average pooling operation; Using the formula W k1 =up(C k )*H k2 ; W k2 =Z2(k-1)*H k3 *σ(Z2(k-1)+C k )。 )。 Z2(k) = W k2 *H k4 Calculate and obtain the multi-data fusion output feature Z2 (k); up() represents the bilinear interpolation operation; σ() represents the sigmoid activation function; * represents the convolution operation; W k1 is the first output value of the process, W k2 Represented as the second output value of the process; In the feature concatenation layer, the multi-data fusion output feature Z1 (K) and the multi-data fusion output feature Z2 (K) are merged and connected to obtain the multi-data feedback feature value of the bipolar electrocoagulation forceps.

[0008] As a preferred technical solution of the present invention, the specific steps of performing abnormal data analysis in the abnormal judgment layer include: Based on the characteristic values ​​of multi-data feedback of bipolar electrocoagulation forceps, reverse clustering analysis was performed to obtain the outliers of multi-data feedback of bipolar electrocoagulation forceps. Perform abnormal identification on outliers in multi-data feedback of bipolar electrocoagulation forceps and obtain abnormal data analysis results; In the abnormal data analysis result, if the abnormal proportion of the outliers in the multi-data feedback of the bipolar electrocoagulation forceps reaches the preset working abnormality threshold, the abnormal data analysis result is abnormal; otherwise, the abnormal data analysis result is non-abnormal; The specific steps for training the anomaly judgment layer include: Collecting several groups of data anomaly judgment training samples; each group of data anomaly judgment training samples contains an abnormal label value and a verified feedback feature value; combining several groups of data anomaly judgment training samples to obtain a data anomaly judgment training set; The data anomaly judgment training set is used for model training to obtain the initial anomaly judgment layer; the initial anomaly judgment layer is evaluated, and if the initial anomaly judgment layer passes the model evaluation, the initial anomaly judgment layer is used as the anomaly judgment layer in the bipolar electrocoagulation forceps data fusion analysis model; otherwise, the data anomaly judgment training set is used to continue model training.

[0009] As a preferred technical solution of the present invention, the real-time feedback management model of the bipolar electrocoagulation forceps includes a feature recognition layer, a feature property mapping layer and a feedback strategy output layer; The feature recognition layer is used to identify the real-time feedback features of the bipolar electrocoagulation forceps and obtain the current bipolar electrocoagulation forceps management strategy; The characteristic mapping layer is used to perform feedback adjustment according to the current bipolar coagulation forceps management strategy and the real-time feedback characteristics of the bipolar coagulation forceps to obtain the real-time management strategy of the bipolar coagulation forceps; The feedback strategy output layer is used to output the real-time management strategy of bipolar electrocoagulation forceps.

[0010] As a preferred technical solution of the present invention, the specific steps of performing feedback adjustment in the characteristic mapping layer include: The current bipolar coagulation forceps management strategy and the bipolar coagulation forceps real-time feedback features are spliced ​​to obtain a comprehensive feature vector for bipolar coagulation forceps management; According to the bipolar coagulation forceps management comprehensive feature vector and the bipolar coagulation forceps management rule mapping, the bipolar coagulation forceps management comprehensive feature vector is mapped to obtain the bipolar coagulation forceps real-time management strategy; The management rule mapping of bipolar electrocoagulation forceps is constructed based on the attention mechanism and swarm optimization algorithm.

[0011] As a preferred technical solution of the present invention, the abnormal signal processing model of the bipolar electrocoagulation forceps includes an abnormal signal recognition layer, an abnormal monitoring and early warning layer, and an early warning signal output layer; The abnormal signal recognition layer is used to extract features of abnormal feedback signals to obtain monitoring abnormal feature signals; The abnormal monitoring and early warning layer is used to classify and identify abnormal monitoring feature signals and obtain corresponding real-time early warning signals; The abnormal monitoring and early warning layer includes a classification signal layer and an early warning analysis layer; the classification signal layer and the early warning analysis layer are trained by a BP neural network model; an LSTM layer is merged into two hidden layers to improve the prediction accuracy, and the hidden layer-LSTM layer-hidden layer is used as the basic structure of the abnormal monitoring and early warning layer; In the classification signal layer, the monitoring abnormal feature signal is classified to obtain the monitoring abnormal feature classification information; In the early warning analysis layer, the classification information of monitoring abnormal features is predicted and analyzed to obtain real-time early warning signals; The warning signal output layer is used to output real-time warning signals.

[0012] A bipolar electrocoagulation forceps is used to execute the above-mentioned bipolar electrocoagulation forceps management method based on multi-data fusion.

[0013] The present invention has the following advantages: 1. The present invention realizes accurate extraction of real-time feedback features and efficient detection of abnormal signals by real-time monitoring of patient physiological parameters, target affected area characteristics and imaging data, and fusion analysis of multi-source data of bipolar coagulation forceps parameters, working environment and forceps tip acquisition signals; combines the real-time feedback management model to intelligently generate management strategies, dynamically optimizes and adjusts the bipolar coagulation forceps, and provides real-time warnings based on the abnormal signal processing model to ensure the safety and stability of equipment operation. This method significantly improves the intelligence level, work efficiency and clinical adaptability of bipolar coagulation forceps, and provides reliable support for high-precision management during surgery.

[0014] 2. The present invention analyzes feedback features in real time through the feature recognition layer, and combines the feature mapping rules constructed based on the attention mechanism and the swarm optimization algorithm to accurately generate a real-time management strategy for bipolar electrocoagulation forceps that meets the current surgical status, thereby significantly improving the management accuracy and intelligence level; in the feature mapping layer, the feedback features and the comprehensive feature vector of the current management strategy are used for feature mapping, so that the device can perceive the dynamic changes of the surgical environment and working status in real time, thereby enhancing the adaptability of the device; the rapid generation and output of the real-time management strategy for bipolar electrocoagulation forceps ensures the real-time response capability of the device in complex surgical scenarios, effectively improving the safety of the surgery and the stability of the operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The present invention is a flowchart of a method for managing bipolar electrocoagulation forceps based on multi-data fusion adopted in an embodiment of the present invention.

[0016] Figure 2 A schematic diagram of the structure of a bipolar electrocoagulation forceps used in an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to enable persons skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0018] A bipolar electrocoagulation forceps management method based on multi-data fusion, see Figure 1 As shown, the following steps are included: Step S1: real-time data monitoring of bipolar electrocoagulation forceps; Acquire real-time monitoring data of patients; real-time monitoring data of patients includes physiological parameter data, characteristic data of target affected area and image data of affected area; At the same time, the real-time working data of the bipolar electrocoagulation forceps is obtained; the real-time working data of the bipolar electrocoagulation forceps includes the parameter data of the bipolar electrocoagulation forceps, the working environment data and the signal data collected by the forceps tip; The real-time monitoring data of patients include physiological parameter data, target affected area characteristic data and affected area image data. The physiological parameter data include vital signs such as heart rate, blood pressure, blood oxygen saturation, respiratory rate, body temperature, electrocardiogram, blood sugar level and blood lactate. The target affected area characteristic data include tissue electrical impedance, tissue water content, temperature distribution and tissue elastic coefficient. The affected area image data include real-time ultrasound images, CT / MRI slice images, infrared thermal images and high-definition camera data. Specific data selection is carried out according to the actual surgical situation. The real-time working data of bipolar electrocoagulation forceps include equipment parameter data, working environment data and forceps tip acquisition signal data. The equipment parameter data covers current, voltage, power output, frequency and working mode. The working environment data involves ambient temperature, humidity, equipment vibration parameters and electromagnetic interference level. The forceps tip acquisition signal data includes contact resistance, temperature sensor signal, tissue feedback signal and tissue status signal, etc. Specific data selection is carried out according to the actual surgical situation. Different data come from different sensors or data acquisition devices, such as monitors, biosensors and laboratory tests; such as tissue impedance meters, ultrasonic elastic imaging devices and infrared temperature sensors; or collected by the built-in control module of the electrocoagulation forceps, or detected and transmitted in real time through the sensor integrated in the forceps tip; According to the patient's real-time monitoring data and the real-time working data of the bipolar coagulation forceps, a real-time data analysis is performed using a bipolar coagulation forceps data fusion analysis model to obtain the real-time feedback characteristics of the bipolar coagulation forceps; wherein, when an abnormal feedback signal is identified in the real-time feedback characteristics of the bipolar coagulation forceps, step S3 is entered for abnormal processing; otherwise, the real-time feedback characteristics of the bipolar coagulation forceps are input into step S2 for analysis; the bipolar coagulation forceps data fusion analysis model realizes real-time and accurate analysis of multi-source data and intelligent feedback adjustment through multi-level data alignment, fusion and feature extraction, which effectively improves the working efficiency, adaptability and safety of the bipolar coagulation forceps; The data fusion analysis model of bipolar electrocoagulation forceps includes data preprocessing layer, multi-data fusion analysis layer, abnormality judgment layer and result output layer; The data preprocessing layer is used to time-align and preprocess the patient real-time monitoring data and the bipolar electrocoagulation forceps real-time working data using the DTW algorithm to obtain preprocessed patient real-time monitoring data and preprocessed bipolar electrocoagulation forceps real-time working data; The multi-data fusion analysis layer is used to perform data fusion analysis on the pre-processed patient real-time monitoring data and the pre-processed bipolar electrocoagulation forceps real-time working data to obtain the multi-data feedback characteristic value of the bipolar electrocoagulation forceps; The abnormal judgment layer is used to perform abnormal data analysis based on the multi-data feedback characteristic values ​​of the bipolar electrocoagulation forceps to obtain the abnormal data analysis result; if the abnormal data analysis result is abnormal, an abnormal feedback signal is output; otherwise, the output is empty; The result output layer is used to output the real-time feedback features of the bipolar coagulation forceps according to the multi-data feedback feature values ​​and abnormal data analysis results of the bipolar coagulation forceps; By introducing a multi-level data processing model, the efficient fusion and analysis of the patient's real-time monitoring data and the real-time working data of the bipolar coagulation forceps are achieved. The DTW algorithm of the data preprocessing layer is used to time align and preprocess the multi-source heterogeneous data to ensure the consistency and integrity of the data; the multi-source data features are deeply extracted and accurately fused through the multi-data fusion analysis layer to obtain the multi-data feedback feature values ​​of the equipment; the feedback features are intelligently detected in combination with the abnormal judgment layer to effectively identify potential problems in the equipment and the surgical process; real-time feedback features are generated in the result output layer for intelligent adjustment of the management strategy of the bipolar coagulation forceps or triggering early warning signals. This method significantly improves the real-time, accuracy and safety of the bipolar coagulation forceps, enhances the intelligence level and clinical adaptability of the equipment, ensures the efficiency and safety of the surgical process, and provides important support for precision medicine; The specific steps of data fusion analysis in the multi-data fusion analysis layer include: In the multi-data fusion analysis layer, there are feature division layer, K convolution kernel layers H k and feature concatenation layer, k=1, 2, ..., K; the specific value of K is set by professional technicians according to actual conditions; The convolution kernel layer H k The convolution kernel in is divided into convolution kernel {H kn}, n=1, 2, ..., 4; convolution kernel {H kn}Equal dimensions between In the feature partitioning layer: Feature extraction is performed on the pre-processed patient real-time monitoring data and the pre-processed bipolar electrocoagulation forceps real-time working data to obtain the patient real-time monitoring data features and the bipolar electrocoagulation forceps real-time working data features; the patient real-time monitoring data features and the bipolar electrocoagulation forceps real-time working data features are spliced ​​and reorganized to obtain multi-data fusion features Z1 (k) and multi-data fusion features Z2 (k); in the feature division layer, k=0 at this time; In the convolution kernel layer H k middle: For the multi-data fusion feature Z1(k-1), the formula Z1(k)=Z1(k-1)*H is used. k1 Perform convolution operation to obtain multi-data fusion output feature Z1(k); Perform multi-resolution convolution transformation on the multi-data fusion feature Z2(k-1) to obtain the multi-data fusion output feature Z2(k); Specific steps: Using formula C k =avgPool(R k )[Z2(k-1)] performs pooling operation to obtain the pooled multi-data fusion output feature C k ; Among them, R k Represents the pooling step size; avgPool[] represents the average pooling operation; Using the formula W k1 =up(C k )*H k2 ; W k2 =Z2(k-1)*H k3 *σ(Z2(k-1)+C k )。 Z2(k)=W k2 *H k4 Calculate and obtain the multi-data fusion output feature Z2 (k); up() represents the bilinear interpolation operation; σ() represents the sigmoid activation function; * represents the convolution operation; W k1 is the first output value of the process, W k2 Represented as the second output value of the process; In the feature concatenation layer, the multi-data fusion output feature Z1 (K) and the multi-data fusion output feature Z2 (K) are merged and connected to obtain the multi-data feedback feature value of the bipolar electrocoagulation forceps; In the feature division layer, independent features are extracted from the pre-processed patient monitoring data and bipolar electrocoagulation forceps working data, and then reorganized into multi-data fusion features through feature splicing, realizing the effective combination of heterogeneous data features; feature division ensures the independence of key information from different data sources in subsequent processing, and provides a basis for multi-resolution convolution; By dividing the convolution kernel into sub-convolution kernels and keeping the dimensions equal, it ensures the extraction of input features from multiple angles and scales. The application of multi-resolution convolution transformation combined with pooling operations can capture the key features of data at different resolutions. Multi-resolution convolution enhances the model's ability to perceive changes in feature space. The nonlinear activation function enhances the expressive power of the model, while the bilinear interpolation operation avoids information loss during data processing, enhances the nonlinear mapping capability of the data fusion process, and assigns weights to features through the attention mechanism to achieve more accurate feature extraction. Through multi-level feature division, multi-resolution convolution, dynamic feature update, and efficient splicing analysis, the accurate fusion and extraction of patient and device data is achieved, with significant advantages such as high efficiency, adaptability, and real-time performance, which significantly improves the intelligent management level and surgical safety of bipolar electrocoagulation forceps. The specific steps for abnormal data analysis in the abnormal judgment layer include: Based on the characteristic values ​​of multi-data feedback of bipolar electrocoagulation forceps, reverse clustering analysis was performed to obtain the outliers of multi-data feedback of bipolar electrocoagulation forceps. Perform abnormal identification on outliers in multi-data feedback of bipolar electrocoagulation forceps and obtain abnormal data analysis results; In the abnormal data analysis result, if the abnormal proportion of the outliers in the multi-data feedback of the bipolar electrocoagulation forceps reaches the preset working abnormality threshold, the abnormal data analysis result is abnormal; otherwise, the abnormal data analysis result is non-abnormal; the preset working abnormality threshold is set by professional technicians according to actual conditions; Based on the reverse clustering analysis of multi-data feedback feature values, outliers in multi-source data can be effectively captured, abnormal equipment status can be accurately identified, and the sensitivity and reliability of abnormal detection can be improved. By setting the abnormal working threshold, abnormal results are triggered only when the proportion of abnormal outliers exceeds the threshold, avoiding overreaction to occasional abnormal data, reducing the false alarm rate, and enhancing stability. The specific steps for training the anomaly judgment layer include: Collecting several groups of data anomaly judgment training samples; each group of data anomaly judgment training samples contains an abnormal label value and a verified feedback feature value; combining several groups of data anomaly judgment training samples to obtain a data anomaly judgment training set; The data anomaly judgment training set is used for model training to obtain the initial anomaly judgment layer; the initial anomaly judgment layer is evaluated, and if the initial anomaly judgment layer passes the model evaluation, the initial anomaly judgment layer is used as the anomaly judgment layer in the bipolar electrocoagulation forceps data fusion analysis model; otherwise, the data anomaly judgment training set is used to continue model training; by constructing a data anomaly judgment training set containing abnormal labels and verified feedback feature values, effective training of the anomaly judgment layer is achieved; combined with the model evaluation mechanism, it is ensured that the performance of the model meets the application requirements, and the accuracy of the anomaly judgment layer is improved through cyclic optimization; the design of the anomaly judgment layer fully considers the diversity and complexity of the real-time feedback data of the bipolar electrocoagulation forceps, and through iterative optimization of training samples and evaluation processes, it has high adaptability to different surgical scenarios and equipment working conditions; by accurately identifying anomalies and generating early warning signals in a timely manner, the real-time performance of equipment management is effectively improved, the safety of the surgical process is guaranteed, and technical support is provided for precise management in complex clinical scenarios; Step S2: real-time management of bipolar electrocoagulation forceps; The real-time feedback features of the bipolar coagulation forceps are input into the real-time feedback management model of the bipolar coagulation forceps for analysis to obtain the real-time management strategy of the bipolar coagulation forceps; the bipolar coagulation forceps are intelligently managed according to the real-time management strategy of the bipolar coagulation forceps; the real-time feedback management model of the bipolar coagulation forceps realizes the efficient linkage between the real-time feedback features and the management strategy through feature recognition, feature mapping based on the attention mechanism and the swarm optimization algorithm, and intelligent strategy output; The real-time feedback management model of bipolar electrocoagulation forceps includes feature recognition layer, feature property mapping layer and feedback strategy output layer; The feature recognition layer is used to identify the real-time feedback features of the bipolar electrocoagulation forceps and obtain the current bipolar electrocoagulation forceps management strategy; The characteristic mapping layer is used to perform feedback adjustment according to the current bipolar coagulation forceps management strategy and the real-time feedback characteristics of the bipolar coagulation forceps to obtain the real-time management strategy of the bipolar coagulation forceps; The feedback strategy output layer is used to output the real-time management strategy of bipolar electrocoagulation forceps; The specific steps of feedback adjustment in the feature mapping layer include: The current bipolar coagulation forceps management strategy and the bipolar coagulation forceps real-time feedback features are spliced ​​to obtain a comprehensive feature vector for bipolar coagulation forceps management; According to the bipolar coagulation forceps management comprehensive feature vector and the bipolar coagulation forceps management rule mapping, the bipolar coagulation forceps management comprehensive feature vector is mapped to obtain the bipolar coagulation forceps real-time management strategy; The management rule mapping of bipolar electrocoagulation forceps is constructed based on the attention mechanism and swarm optimization algorithm; By analyzing the feedback features in real time through the feature recognition layer, combined with the feature mapping rules built based on the attention mechanism and the group optimization algorithm, the real-time management strategy of the bipolar electrocoagulation forceps that conforms to the current surgical status is accurately generated, which significantly improves the management accuracy and intelligence level; in the feature mapping layer, the feedback features and the comprehensive feature vector of the current management strategy are used for feature mapping, so that the device can perceive the dynamic changes of the surgical environment and working status in real time, and enhance the adaptability of the device; the rapid generation and output of the real-time management strategy of the bipolar electrocoagulation forceps ensures the real-time response capability of the device in complex surgical scenarios, effectively improving the safety of the surgery and the stability of the operation; Specific method for mapping bipolar electrocoagulation forceps management rules: collect real-time feedback features and historical management strategy data of bipolar electrocoagulation forceps, including equipment working parameters, patient monitoring data, and optimal management strategy labels under corresponding surgical scenarios, standardize and normalize these data to ensure the numerical consistency of different feature dimensions, and clean abnormal data to ensure the quality of training samples; Design an attention mechanism to identify the key features that have the greatest impact on the management strategy from the real-time feedback features of the bipolar electrocoagulation forceps, input the real-time feedback features into the attention module, and highlight the important features by calculating the weight distribution; perform a weighted combination of the features according to the attention weights to obtain the aggregated key feature vector; The swarm optimization algorithm is used to preliminarily model the bipolar electrocoagulation forceps management rules and generate multiple sets of candidate rule mapping models; a set of initial candidate rule mappings is randomly generated in the search space and represented as particles; according to the performance objectives of the bipolar electrocoagulation forceps management strategy, such as optimal power output, minimum tissue damage, etc., a fitness function is designed to evaluate the effect of each set of rule mappings; in particle swarm optimization, the position and velocity of each particle are updated using a formula, and in the genetic algorithm, operations such as crossover and mutation are used to generate the next generation of rules, and the optimal rule is selected through the fitness function; the preliminary rule mapping model generated by the swarm optimization algorithm is combined with the key feature vectors extracted by the attention mechanism to construct a comprehensive management rule mapping; a neural network is used to fit the preliminary rules, and attention weights are introduced as adjustment factors during the training process to optimize the prediction accuracy of rule mappings; the performance of the rule mapping model is evaluated using a cross-validation method, including accuracy, stability, and robustness. For the parts that perform poorly in the evaluation, the model is optimized by re-adjusting the attention weights or optimizing parameters until the application requirements are met; The above steps only provide a construction method. The actual construction rules are set by professional technicians according to actual conditions. Step S3: abnormal monitoring and early warning; When an abnormal feedback signal is received, an analysis is performed based on the real-time feedback characteristics of the bipolar electrocoagulation forceps and the abnormal signal processing model of the bipolar electrocoagulation forceps to obtain a real-time warning signal; follow-up measures are taken according to the real-time warning signal; The abnormal signal processing model of bipolar electrocoagulation forceps includes an abnormal signal recognition layer, an abnormal monitoring and early warning layer, and an early warning signal output layer; The abnormal signal recognition layer is used to extract features of abnormal feedback signals to obtain monitoring abnormal feature signals; The abnormal monitoring and early warning layer is used to classify and identify abnormal monitoring feature signals and obtain corresponding real-time early warning signals; The abnormal monitoring and early warning layer includes a classification signal layer and an early warning analysis layer; the classification signal layer and the early warning analysis layer are trained by a BP neural network model; an LSTM layer is merged into two hidden layers to improve the prediction accuracy, and the hidden layer-LSTM layer-hidden layer is used as the basic structure of the abnormal monitoring and early warning layer; In the classification signal layer, the monitoring abnormal feature signal is classified to obtain the monitoring abnormal feature classification information; In the early warning analysis layer, the classification information of monitoring abnormal features is predicted and analyzed to obtain real-time early warning signals; The warning signal output layer is used to output real-time warning signals; The abnormal signal recognition layer performs deep feature extraction on the received abnormal feedback signals, effectively captures the key feature signals under abnormal conditions, lays a high-quality data foundation for subsequent classification and prediction, and improves the accuracy of abnormal detection; the abnormal monitoring and early warning layer adopts a fusion model based on BP neural network and LSTM, fully utilizing the nonlinear feature extraction capability of BP neural network and the prediction advantage of LSTM for time series data. Through the hidden layer-LSTM layer-hidden layer structural design, it effectively improves the accuracy of abnormal classification and the reliability of early warning signals; the classification signal layer quickly classifies abnormal signals, and combined with the real-time prediction function of the early warning analysis layer, it can quickly generate corresponding early warning signals, realize dynamic monitoring and real-time early warning of equipment abnormalities, and meet the needs of high real-time response during surgery; the modular design of the abnormal signal recognition layer, abnormal monitoring and early warning layer, and early warning signal output layer makes the abnormal signal processing model highly flexible and extensible, and the classification standard and prediction model can be easily adjusted according to different application scenarios; In the actual implementation process, for example, in a complex brain tumor resection operation, the surgeon uses bipolar electrocoagulation forceps to perform tissue cutting and coagulation operations. The bipolar electrocoagulation forceps monitor the patient's physiological parameters (such as intracranial pressure and brain swelling) and the tissue characteristics of the target affected area (such as the texture and elasticity of the tumor site) in real time; through dynamic monitoring of the texture and elasticity of the target area, surgeons can more accurately evaluate the physical properties of tissues in different parts, so as to select the best tumor resection approach and formulate more scientific surgical paths and strategies to minimize damage to normal brain tissue; in addition, the device also collects working parameters and forceps tip feedback signals to generate multimodal data streams to support real-time decision-making; during the operation, when abnormal signals appear in the device feedback characteristics, such as changes in local resistance due to heterogeneity of tumor tissue density or texture, which in turn causes differences in thermal effects during the operation, the abnormal signals are automatically identified and input into the abnormal signal processing unit. In the management model, the abnormal signal recognition layer quickly extracts key features, and the abnormal monitoring and early warning layer predicts potential abnormal equipment status through LSTM and BP neural network structures to generate real-time early warning signals. At this time, the early warning signal output layer sends risk prompts to the surgical display screen and alarm system to remind doctors to pay attention to equipment parameter adjustments or suspend operations to prevent thermal damage to normal brain tissue. At the same time, the real-time feedback management model of bipolar electrocoagulation forceps dynamically generates management strategies based on the current equipment status and patient feedback characteristics. For example, when the temperature of the surgical area is detected to be increased, the device will automatically reduce the output power or adjust the voltage to avoid overheating or carbonization of brain tissue, thereby protecting the normal structure of key functional areas. Throughout the process, the device relies on a multi-data fusion analysis model to integrate patient data and equipment data, realize intelligent dynamic regulation, provide doctors with safe and efficient operation support, and effectively improve the accuracy of brain tumor surgery and patient safety.

[0019] A bipolar electrocoagulation forceps, see Figure 2 As shown, including: The tweezers tip and tweezers rod are precision machined to ensure that they have a shape that meets surgical needs. The tweezers handle is formed by laser cutting and CNC processing. The tweezers rod and handle are connected together by welding. The sensor is embedded in the slot of the tweezers rod and tweezers handle. The sensor and the connecting wire are fixed to the tweezers rod by heat shrink tubing. The tweezers handle and sensor connecting wire are wrapped by silicone encapsulation. Wires are welded to the tail ends of the left and right tweezers and the left and right tweezers are aligned and fixed with glue using tooling. A detachable tweezer tip design has been added to the traditional electrocoagulation tweezers, and the practicality and maintenance convenience of the equipment have been improved by optimizing the structure. Patients can choose different types of tweezer tips according to different surgical needs, such as pointed, flat, elbow, etc., to meet the requirements of various tissue types and surgical scenarios. After surgery, the tweezers tips can be sterilized or replaced separately at high temperature to reduce maintenance costs. At the same time, the tweezers tips are equipped with a built-in sensor module for acquiring data.

[0020] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all such improvements and changes should fall within the scope of protection of the appended claims of the present invention. Parts not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims

1. A bipolar electrocoagulation forceps management method based on multi-data fusion, characterized in that: The following steps are involved: Step S1: real-time data monitoring of bipolar electrocoagulation forceps; Obtain real-time patient monitoring data; The patient's real-time monitoring data includes physiological parameter data, target affected area characteristic data and affected area image data; At the same time, the real-time working data of the bipolar electrocoagulation forceps is obtained; the real-time working data of the bipolar electrocoagulation forceps includes the parameter data of the bipolar electrocoagulation forceps, the working environment data and the signal data collected by the forceps tip; According to the patient's real-time monitoring data and the real-time working data of the bipolar coagulation forceps, a real-time data analysis is performed using a bipolar coagulation forceps data fusion analysis model to obtain the real-time feedback characteristics of the bipolar coagulation forceps; wherein, when an abnormal feedback signal is identified in the real-time feedback characteristics of the bipolar coagulation forceps, step S3 is entered for abnormal processing; otherwise, the real-time feedback characteristics of the bipolar coagulation forceps are input into step S2 for analysis; the bipolar coagulation forceps data fusion analysis model realizes real-time and accurate analysis of multi-source data and intelligent feedback adjustment through multi-level data alignment, fusion and feature extraction, which effectively improves the working efficiency, adaptability and safety of the bipolar coagulation forceps; Step S2: real-time management of bipolar electrocoagulation forceps; The real-time feedback features of the bipolar coagulation forceps are input into the real-time feedback management model of the bipolar coagulation forceps for analysis to obtain the real-time management strategy of the bipolar coagulation forceps; the bipolar coagulation forceps are intelligently managed according to the real-time management strategy of the bipolar coagulation forceps; the real-time feedback management model of the bipolar coagulation forceps realizes the efficient linkage between the real-time feedback features and the management strategy through feature recognition, feature mapping based on the attention mechanism and the swarm optimization algorithm, and intelligent strategy output; Step S3: abnormal monitoring and early warning; When an abnormal feedback signal is received, an analysis is performed based on the real-time feedback characteristics of the bipolar coagulation forceps and the abnormal signal processing model of the bipolar coagulation forceps to obtain a real-time warning signal; subsequent measures are taken according to the real-time warning signal.

2. A bipolar electrocoagulation forceps management method based on multi-data fusion according to claim 1, characterized in that: The data fusion analysis model of bipolar electrocoagulation forceps includes data preprocessing layer, multi-data fusion analysis layer, abnormality judgment layer and result output layer; The data preprocessing layer is used to time-align and preprocess the patient real-time monitoring data and the bipolar electrocoagulation forceps real-time working data using the DTW algorithm to obtain preprocessed patient real-time monitoring data and preprocessed bipolar electrocoagulation forceps real-time working data; The multi-data fusion analysis layer is used to perform data fusion analysis on the pre-processed patient real-time monitoring data and the pre-processed bipolar electrocoagulation forceps real-time working data to obtain the multi-data feedback characteristic value of the bipolar electrocoagulation forceps; The abnormal judgment layer is used to perform abnormal data analysis based on the multi-data feedback characteristic values ​​of the bipolar electrocoagulation forceps to obtain abnormal data analysis results; if the abnormal data analysis result is abnormal, an abnormal feedback signal is output; Otherwise, the output is empty; The result output layer is used to output the real-time feedback features of the bipolar coagulation forceps according to the multi-data feedback feature values ​​and abnormal data analysis results of the bipolar coagulation forceps.

3. A bipolar electrocoagulation forceps management method based on multi-data fusion according to claim 2, characterized in that: The specific steps of data fusion analysis in the multi-data fusion analysis layer include: In the multi-data fusion analysis layer, there are feature division layer, K convolution kernel layers H k and feature concatenation layer, k=1,2,…,K; The convolution kernel layer H k The convolution kernel in is divided into convolution kernel {H kn }, n=1, 2, ..., 4; convolution kernel {H kn }Equal dimensions between In the feature partitioning layer: Feature extraction is performed on the pre-processed patient real-time monitoring data and the pre-processed bipolar electrocoagulation forceps real-time working data to obtain the patient real-time monitoring data features and the bipolar electrocoagulation forceps real-time working data features; the patient real-time monitoring data features and the bipolar electrocoagulation forceps real-time working data features are spliced ​​and reorganized to obtain multi-data fusion features Z1 (k) and multi-data fusion features Z2 (k); in the feature division layer, k=0 at this time; In the convolution kernel layer H k middle: For the multi-data fusion feature Z1(k-1), the formula Z1(k)=Z1(k-1)*H is used. k1 Perform convolution operation to obtain multi-data fusion output feature Z1(k); Perform multi-resolution convolution transformation on the multi-data fusion feature Z2(k-1) to obtain the multi-data fusion output feature Z2(k); Specific steps: Using formula C k =avgPool(R k )[Z2(k-1)] performs pooling operation to obtain the pooled multi-data fusion output feature C k ; Among them, R k Represents the pooling step size; avgPool[] represents the average pooling operation; Using the formula W k1 =up(C k )*H k2 ; W k2 =Z2(k-1)*H k3 *σ(Z2(k-1)+C k )。 )。 Z2(k) = W k2 *H k4 Calculate and obtain the multi-data fusion output feature Z2 (k); up() represents the bilinear interpolation operation; σ() represents the sigmoid activation function; * represents the convolution operation; W k1 is the first output value of the process, W k2 Represented as the second output value of the process; In the feature concatenation layer, the multi-data fusion output feature Z1 (K) and the multi-data fusion output feature Z2 (K) are merged and connected to obtain the multi-data feedback feature value of the bipolar electrocoagulation forceps.

4. A bipolar electrocoagulation forceps management method based on multi-data fusion according to claim 3, characterized in that: The specific steps for abnormal data analysis in the abnormal judgment layer include: Based on the characteristic values ​​of multi-data feedback of bipolar electrocoagulation forceps, reverse clustering analysis was performed to obtain the outliers of multi-data feedback of bipolar electrocoagulation forceps. Perform abnormal identification on outliers in multi-data feedback of bipolar electrocoagulation forceps and obtain abnormal data analysis results; In the abnormal data analysis result, if the abnormal proportion of the outliers in the multi-data feedback of the bipolar electrocoagulation forceps reaches the preset working abnormality threshold, the abnormal data analysis result is abnormal; otherwise, the abnormal data analysis result is non-abnormal; The specific steps for training the anomaly judgment layer include: Collecting several groups of data anomaly judgment training samples; each group of data anomaly judgment training samples contains an abnormal label value and a verified feedback feature value; combining several groups of data anomaly judgment training samples to obtain a data anomaly judgment training set; The data anomaly judgment training set is used for model training to obtain the initial anomaly judgment layer; the initial anomaly judgment layer is evaluated, and if the initial anomaly judgment layer passes the model evaluation, the initial anomaly judgment layer is used as the anomaly judgment layer in the bipolar electrocoagulation forceps data fusion analysis model; otherwise, the data anomaly judgment training set is used to continue model training.

5. A bipolar electrocoagulation forceps management method based on multi-data fusion according to claim 4, characterized in that: The real-time feedback management model of bipolar electrocoagulation forceps includes feature recognition layer, feature property mapping layer and feedback strategy output layer; The feature recognition layer is used to identify the real-time feedback features of the bipolar electrocoagulation forceps and obtain the current bipolar electrocoagulation forceps management strategy; The characteristic mapping layer is used to perform feedback adjustment according to the current bipolar coagulation forceps management strategy and the real-time feedback characteristics of the bipolar coagulation forceps to obtain the real-time management strategy of the bipolar coagulation forceps; The feedback strategy output layer is used to output the real-time management strategy of bipolar electrocoagulation forceps.

6. A bipolar electrocoagulation forceps management method based on multi-data fusion according to claim 5, characterized in that: The specific steps of feedback adjustment in the feature mapping layer include: The current bipolar coagulation forceps management strategy and the bipolar coagulation forceps real-time feedback features are spliced ​​to obtain a comprehensive feature vector for bipolar coagulation forceps management; According to the bipolar coagulation forceps management comprehensive feature vector and the bipolar coagulation forceps management rule mapping, the bipolar coagulation forceps management comprehensive feature vector is mapped to obtain the bipolar coagulation forceps real-time management strategy; The management rule mapping of bipolar electrocoagulation forceps is constructed based on the attention mechanism and swarm optimization algorithm.

7. A bipolar electrocoagulation forceps management method based on multi-data fusion according to claim 6, characterized in that: The abnormal signal processing model of bipolar electrocoagulation forceps includes an abnormal signal recognition layer, an abnormal monitoring and early warning layer, and an early warning signal output layer; The abnormal signal recognition layer is used to extract features of abnormal feedback signals to obtain monitoring abnormal feature signals; The abnormal monitoring and early warning layer is used to classify and identify abnormal monitoring feature signals and obtain corresponding real-time early warning signals; The abnormal monitoring and early warning layer includes a classification signal layer and an early warning analysis layer; the classification signal layer and the early warning analysis layer are trained by a BP neural network model; an LSTM layer is merged into two hidden layers to improve the prediction accuracy, and the hidden layer-LSTM layer-hidden layer is used as the basic structure of the abnormal monitoring and early warning layer; In the classification signal layer, the monitoring abnormal feature signal is classified to obtain the monitoring abnormal feature classification information; In the early warning analysis layer, the classification information of monitoring abnormal features is predicted and analyzed to obtain real-time early warning signals; The warning signal output layer is used to output real-time warning signals.

8. A bipolar electrocoagulation forceps, characterized in that: The system applies a bipolar electrocoagulation forceps management method based on multi-data fusion as described in any one of claims 1 to 7 above.

Images