Soft tissue puncture path planning and monitoring method and system based on electrical impedance tomography

Abstract

The invention discloses a soft tissue puncture path planning and monitoring method and system based on electrical impedance tomography. The method comprises the steps that before puncture, multi-modal soft tissue information is obtained through multi-frequency-domain electrical impedance imaging, ultrasonic imaging and nuclear magnetic resonance imaging; establishing an image reconstruction model, and fusing multi-modal information to improve the imaging resolution; performing preoperative path planning modeling, defining a state space and an action space, and designing global and local reward function optimization path strategies; a mechanical sensor at the front end of the puncture needle is used for monitoring path resistance in real time and dynamically adjusting a path during operation; and the optimal route design is achieved through iterative training. The system comprises a data acquisition module, an image reconstruction module, a route planning module, an intraoperative monitoring module and an abnormity alarm module, the tissue damage rate can be effectively reduced, the operation time is shortened, and precise and intelligent technical support is provided for the minimally invasive interventional operation.

Description

Technical Field

[0001] The present invention relates to the fields of medicine and computer technology, and particularly relates to a method and system for soft tissue puncture path planning and monitoring based on electrical impedance tomography. Background Art

[0002] Electrical impedance tomography (EIT) is a functional imaging technology based on the conductivity differences of biological tissues. Its basic principle is to inject a weak alternating current through a surface electrode array, measure the boundary voltage changes, and reconstruct the conductivity distribution inside the tissue by combining inverse problem solving algorithms. Since there are significant differences in the conductivities of different soft tissues, EIT can provide a non-invasive, real-time, and non-radiative imaging method, which has important application values in the fields of medical monitoring and intraoperative navigation.

[0003] In recent years, EIT has been introduced into the field of soft tissue puncture path planning to assist surgical navigation. However, existing EIT-based puncture path planning methods still have key problems such as low imaging resolution, poor path planning flexibility, and insufficient intelligence; therefore, there is an urgent need to develop a high-precision, adaptive, and intelligent EIT-guided puncture path planning method to improve the surgical success rate and patient safety. Summary of the Invention

[0004] Aiming at the deficiencies of existing detection methods, a method for soft tissue puncture path planning and monitoring based on electrical impedance tomography is proposed, and a system for soft tissue puncture path planning and monitoring based on electrical impedance tomography is built.

[0005] The method for soft tissue puncture path planning and monitoring based on electrical impedance tomography is characterized in that:

[0006] Step S1: Before puncture, obtain soft tissue information in different frequency domains through multi-frequency electrical impedance tomography, and at the same time obtain richer structural information through ultrasonic imaging and magnetic resonance imaging.

[0007] Step S2: Establish an image reconstruction model, fuse multi-frequency electrical impedance tomography information, ultrasonic imaging information, and magnetic resonance imaging information, train a prior auxiliary network, and improve the resolution with the help of supplementary information.

[0008] Step S3: Preoperative path planning modeling, the preoperative path planning model is trained based on the idea of reinforcement learning. The needle tip position and the relative position with the target point during the puncture process are used as state information, and a suitable reward function is defined. For example, a positive reward is given when the needle tip approaches the target point or reaches successfully, and a negative reward is given when hitting an obstacle or deviating from the path. Through continuous interaction with the environment and trial-and-error learning, the intelligent agent can gradually master the optimal action strategy in different states;

[0009] Step S4: Install a mechanical sensor at the front end of the puncture needle, monitor the resistance of the puncture path in real time during the operation, analyze the state of the surrounding soft tissues through the tip sensor, and dynamically adjust the path.

[0010] Introduce the tip sensor data N for real-time monitoring of the rationality of the path during the operation and abnormal handling; the local reward function is optimized as:

[0011] Localr = Localr + R_abnormality_of_resistance

[0012] R_abnormality_of_resistance = -d·(Ft - Fsafe)

[0013] This value measures that the resistance Ft at the current moment t exceeds the expected safety threshold Fsafe, and a penalty is imposed according to the over-limit ratio. d is the penalty weight of the resistance abnormality term;

[0014] Step S5: Iteratively train Step S3 - Step S4 until the model converges, that is, the optimal route design is achieved;

[0015] The present invention also relates to a soft tissue puncture path planning and monitoring system based on electrical impedance tomography. The system is used to execute the soft tissue puncture path planning and monitoring method based on electrical impedance tomography as described above. The system includes:

[0016] The present invention also relates to a soft tissue puncture path planning and monitoring system based on electrical impedance tomography. The system is used to execute the soft tissue puncture path planning and monitoring method based on electrical impedance tomography as described above. The system includes:

[0017] A data acquisition module for injecting multi-frequency alternating current through a surface electrode array to obtain conductivity distribution data in different frequency domains; at the same time, collecting ultrasonic imaging data and nuclear magnetic resonance imaging data to provide a multi-modal information basis for subsequent path planning, and ensuring the spatio-temporal consistency and usability of each modal data through data preprocessing;

[0018] An image reconstruction module for constructing a priori auxiliary network, extracting key anatomical structure features such as blood vessels and tumors based on ultrasonic and nuclear magnetic resonance data, and generating a probability map matrix; at the same time, guiding the EIT reconstruction network to fuse multi-frequency domain electrical impedance tomography information with the a priori feature matrix, and using weighted and correlation calculation strategies to improve the resolution of electrical impedance tomography and the ability to identify microstructures;

[0019] The route planning module is used to mark high-risk areas and target tissues based on the image reconstruction results through threshold segmentation and morphological operations; construct a reinforcement learning model, define a state space including impedance features, needle tip position, etc. and an action space including forward direction angle and propulsion force; design global and local reward functions, combine factors such as patient age, target proximity, and distance from dangerous areas, and optimize through iterative optimization of the Value function to plan a safe and efficient puncture path;

[0020] The intraoperative monitoring module is used to collect resistance, resistance change rate, and lateral force data in real time using the mechanical sensor at the front end of the puncture needle; integrate the sensor data into the local reward function to dynamically adjust the puncture path; analyze the data trend, predict potential risks, and achieve real-time dynamic monitoring and path correction during the puncture process;

[0021] The abnormal alarm module is used to monitor abnormal conditions such as resistance exceeding the limit, path deviation, and sensor failure during the puncture process in real time; when an abnormality occurs, issue an alarm through sound, light, pop-up windows, etc., automatically pause the puncture operation or perform safety protection actions, and provide treatment suggestions for medical staff to ensure the safety of the operation;

[0022] The present invention also discloses a non-volatile storage medium, which is characterized in that the non-volatile storage medium includes a stored program, wherein the program controls the device where the non-volatile storage medium is located to execute the above method when running.

[0023] The present invention also discloses an electronic device, which is characterized in that it includes a processor and a memory; the memory stores computer-readable instructions, and the processor is used to run the computer-readable instructions, wherein the computer-readable instructions execute the above method when running.

[0024] Beneficial effects

[0025] The soft tissue puncture path planning and monitoring system and method based on electrical impedance tomography of the present invention fuse multi-modal imaging of electrical impedance, ultrasound, and nuclear magnetic resonance, use a priori auxiliary network and optimization algorithm to improve imaging resolution, construct an intelligent path planning model by combining reinforcement learning, balance efficiency and safety with global and local reward functions, and introduce a mechanical sensor to achieve intraoperative dynamic monitoring and path correction. The system takes into account both high efficiency and strategy adjustment ability. Compared with traditional methods, it can significantly reduce the tissue damage rate and shorten the operation time, and is applicable to a variety of minimally invasive interventional surgeries, providing standardized and intelligent technical support for precision medicine. Brief description of the drawings

[0026] Figure 1 It is the system architecture diagram of the present invention. Detailed implementation manners

[0027] Example 1

[0028] A method for soft tissue puncture path planning and monitoring based on electrical impedance tomography, comprising the following steps:

[0029] Step S1: Before puncture, obtain soft tissue information in different frequency domains through multi-frequency electrical impedance tomography, and at the same time obtain richer structural information through ultrasonic imaging and magnetic resonance imaging.

[0030] Step S11: Inject multi-frequency alternating current through a body surface electrode array to obtain the conductivity distribution in different frequency domains for better differentiation of different tissues, and output the imaging corresponding to the frequency domain as E_f, where f is the frequency;

[0031] Step S12: Obtain ultrasonic imaging data for preoperative supplementation of tissue structure information, denoted as U;

[0032] Step S13: Obtain magnetic resonance imaging data for preoperative supplementation of high-resolution structural and functional information, denoted as M;

[0033] Step S2: Establish an image reconstruction model, fuse multi-frequency electrical impedance tomography information, ultrasonic imaging information and magnetic resonance imaging information, and improve the resolution with the help of supplementary information.

[0034] Step S21: Combine ultrasonic data and magnetic resonance data to train a prior auxiliary network to extract tissue features of key anatomical structures such as blood vessels and tumors as prior constraints for electrical impedance tomography reconstruction; the prior auxiliary network uses an encoder architecture, with the input being ultrasonic imaging data U and magnetic resonance imaging data M, and the output predicting the features of key parts, denoted as the probability map matrix MP, with a range of [0,1], where approaching 0 represents the background and approaching 1 represents key structures such as blood vessels or tumors. The calculation formula is as follows:

[0035] MP = sigmoid(NETWORK(M,U))

[0036] Among them, NETWORK represents the encoder layer, sigmoid represents the activation function. After the ultrasonic imaging data U and magnetic resonance imaging data M pass through the encoder layer and are activated, the predicted probability map MP of the key part is obtained.

[0037] The purpose of modeling the loss function is to make the distribution of the predicted value MP close to the true value MP'. The formula is as follows:

[0038]

[0039] Among them, MP represents the feature data predicted by the model, MP' represents the true feature label, which only has 0 and 1 values, · represents the dot product operation, || represents the summation operation, and μ represents the balance parameter.

[0040] Step S22: Guide the optimization of the EIT reconstruction network, intelligently fuse the multi-frequency domain electrical impedance tomography information E_f and the prior key part feature matrix MP, use the prior probability map and multi-frequency domain information to improve the resolution and enhance the recognition ability of tiny structures; considering the sparsity of the data, the fusion formula is designed as follows:

[0041] E = MEAN(E_f)

[0042] where E represents the electrical impedance tomography data after fusing the multi-frequency domain, and MEAN represents the mean operation;

[0043] Q = Wq·E

[0044] where Q represents the mapped electrical impedance tomography data, and Wq is the weight matrix;

[0045] K = (Wk·MP) T

[0046] where K represents the prior data MP after mapping, which is used to calculate the correlation. For convenient calculation, it is transformed using the T transpose operation, and Wk is the corresponding weight matrix;

[0047] V = Wv·E_f

[0048] where V is used to calculate the output after correlation weighting, and Wv is the corresponding weight matrix;

[0049]

[0050] where A represents the correlation sparse matrix, i, j represent the corresponding positions in the matrix, exp represents the exponential operation, l represents only taking the top-k elements, k is a manually set value, and d represents the scaling dimension; the correlation sparse matrix can pay more attention to the key regions;

[0051] E’ = E + softmax(A)·V

[0052] where E’ represents the optimized electrical impedance tomography features of the output. softmax is the normalization function, which adds the original information E and the feature map softmax(A)·V after calculating the correlation to obtain the optimized electrical impedance tomography features E’;

[0053] Step S3: Preoperative path planning modeling. The preoperative path planning model trains the model based on the idea of reinforcement learning, takes the needle tip position and the relative position with the target point during the puncture process as state information, and defines an appropriate reward function. For example, a positive reward is given when the needle tip approaches the target point or reaches it smoothly, and a negative reward is given when it collides with an obstacle or deviates from the path. By continuously interacting with the environment and trial-and-error learning, the intelligent agent can gradually master the optimal action strategy in different states;

[0054] Step S31: Based on the optimized electrical impedance tomography feature E, detect and label key regions through threshold segmentation and morphological operations. The key regions are divided into high-risk regions (such as blood vessels, nerves) and target tissues (such as tumors);

[0055] Step S32: Path planning modeling

[0056] Modeling objective: Complete the optimal path planning

[0057] Define the action space A: {advancing direction angle, advancing force},

[0058] Define the state space S: {electrical impedance tomography feature E’, needle tip position, high-risk regions and target tissues}

[0059] Global reward function design

[0060] Globalr = R target proximity + R dangerous region penalty + R age coefficient. The global reward function is used to guide the macroscopic obstacle avoidance of the puncture path and efficiently approach the target. The target proximity encourages quickly approaching the target tissue. The penalty for approaching the dangerous region punishes the path for approaching high-risk regions (such as blood vessels, nerves) to ensure a safe distance; the age coefficient adjusts the strategy according to the patient's age;

[0061] R target proximity = -a·||p_cur - p_obj||2

[0062] R target proximity encourages quickly approaching the target tissue, calculated by the Euclidean distance from the needle tip to the target tissue. p_cur represents the needle tip position, p_obj represents the target position, ||||2 represents the Euclidean distance, and a is the weight of this term;

[0063] R dangerous region penalty = -ρ∑ i∈高危区域 ||p_cur - p_danger_i||2

[0064] R dangerous region penalty punishes the path for approaching high-risk regions. i represents the high-risk region number, p_danger_i represents the position of the i-th high-risk region, ||||2 represents the Euclidean distance, and ρ represents the penalty coefficient;

[0065] R age coefficient = -c·|age - 30|

[0066] R age coefficient adjusts the advancing strategy according to the patient's age. age represents the patient's age, c represents the weight of the age term, || is the absolute value. When the patient is an elderly person or a child, the advancing force needs to be reduced;

[0067] Local reward function design;

[0068] Localr = R path smoothness

[0069] The local reward function is used for refined design to balance path efficiency and safety. The R path smoothness encourages a gentle change in the puncture direction of adjacent actions, reducing the risk of tissue tearing caused by sharp turns;

[0070] R path smoothness = -τ∑ i∈动作序列 |θcur_i - θpre_i|

[0071] The R path smoothness encourages a gentle change in the puncture direction of adjacent action sequences. i represents a certain action, θcur_i represents the forward direction angle at the i-th time, θpre_i represents the previous forward direction angle at the i-th time, || belongs to the absolute value operation, and τ represents the weight of smoothness;

[0072] The reward function is the sum of the global and local reward functions;

[0073] R = Globalr + Localr

[0074] The Value function update strategy is as follows:

[0075] V(s t ,a t ) ← V(s t ,a t ) + [R + σmax a' V(s t+1 ,a') - V(s t ,a t )]

[0076] The Value function V is the key to modeling learning. t represents the moment t, s t represents the state at the t-th moment, a t represents the action taken at the t-th moment, R is the above reward function, σ is the discount factor, max a' V(s t+1 ,a') represents selecting an action a’ such that the Value function corresponding to the next moment state s t+1 is the largest;

[0077] Step S4: Install a mechanical sensor at the front end of the puncture needle to monitor the puncture path resistance in real time during the operation. Analyze the state of the surrounding soft tissues through the tip sensor and dynamically adjust the path.

[0078] Introduce the tip sensor data N for real-time monitoring of the path rationality and abnormal handling during the operation; The intraoperative local reward function is optimized as:

[0079] Localr = Localr + R resistance abnormality

[0080] R resistance abnormality = -d·(Ft - Fsafe)

[0081] The abnormal resistance degree R measures that the resistance Ft at the current moment t exceeds the expected safety threshold Fsafe, and a penalty is imposed according to the over-limit ratio. d is the penalty weight of the abnormal resistance term;

[0082] Step S5: Iteratively train steps S3 - S4 until the model converges, that is, the optimal route design is achieved;

[0083] Embodiment 2

[0084] A soft tissue puncture path planning and monitoring system based on electrical impedance tomography, which is used to execute the soft tissue puncture path planning and monitoring method based on electrical impedance tomography as described above, such as Figure 1 , characterized in that:

[0085] A data acquisition module, which is used to inject multi-frequency alternating current through a body surface electrode array to obtain conductivity distribution data in different frequency domains; at the same time, collect ultrasonic imaging data and nuclear magnetic resonance imaging data to provide a multi-modal information basis for subsequent path planning, and ensure the spatio-temporal consistency and usability of each modal data through data preprocessing;

[0086] The data acquisition module further includes:

[0087] An impedance data acquisition sub-module: used to inject multi-frequency alternating current into the human body through a body surface electrode array, measure the boundary voltage change, obtain conductivity distribution data in different frequency domains, and reflect tissue differences from the electrophysiological characteristic level;

[0088] An ultrasonic data acquisition sub-module: using the reflection and refraction characteristics of ultrasonic waves propagating in human tissues, collect high-frame-rate two-dimensional tomographic image data, and quickly provide morphological structure information of soft tissues;

[0089] A nuclear magnetic resonance data acquisition sub-module: relying on a strong magnetic field and radio frequency pulses, obtain high-resolution three-dimensional image data, and provide rich tissue anatomical structure and function information;

[0090] A data preprocessing sub-module: perform operations such as time synchronization, spatial alignment, noise reduction, and format conversion on the collected multi-modal data to ensure the consistency and usability of the data.

[0091] An image reconstruction module, which is used to construct a prior assistance network, extract key anatomical structure features such as blood vessels and tumors based on ultrasonic and nuclear magnetic resonance data, and generate a probability map matrix; at the same time, guide the EIT reconstruction network, fuse the multi-frequency domain electrical impedance tomography information with the prior feature matrix, and use weighted and correlation calculation strategies to improve the resolution of electrical impedance tomography and the ability to identify minute structures;

[0092] The image reconstruction module further includes: a prior feature extraction sub-module and an EIT reconstruction optimization sub-module;

[0093] The prior feature extraction sub-module constructs a prior auxiliary network based on ultrasonic and nuclear magnetic resonance data, extracts key anatomical structure features such as blood vessels and tumors, generates a probability map matrix, and provides prior constraints for electrical impedance tomography reconstruction;

[0094] The EIT reconstruction optimization sub-module fuses multi-frequency electrical impedance tomography information with the prior feature matrix, and uses strategies such as weighting and correlation calculation to optimize the electrical impedance tomography features, improving the image resolution and the ability to identify tiny structures;

[0095] The route planning module is used to mark high-risk areas and target tissues based on the image reconstruction results through threshold segmentation and morphological operations; construct a reinforcement learning model, define a state space including impedance features, needle tip position, etc. and an action space including forward direction angle and propulsion force; design global and local reward functions, and combine factors such as patient age, target proximity, and distance from dangerous areas, and optimize through Value function iteration to plan a safe and efficient puncture path;

[0096] The route planning module further includes: a key area marking sub-module and a reinforcement learning modeling sub-module;

[0097] The key area marking sub-module detects and marks high-risk areas and target tissues based on the optimized electrical impedance tomography features through threshold segmentation and morphological operations;

[0098] The reinforcement learning modeling sub-module is used to construct a reinforcement learning model, define a state space including impedance features, needle tip position, etc., and an action space including forward direction angle and propulsion force; design global and local reward functions, and combine multiple factors, and optimize through Value function iteration to plan a safe and efficient puncture path;

[0099] The intraoperative monitoring module is used to collect resistance, resistance change rate, and lateral force data in real time using the mechanical sensor at the front end of the puncture needle; integrate the sensor data into the local reward function to dynamically adjust the puncture path; analyze the data trend, predict potential risks, and achieve real-time dynamic monitoring and path correction of the puncture process;

[0100] The intraoperative monitoring module further includes: a mechanical data collection sub-module and a path dynamic adjustment sub-module;

[0101] The mechanical data collection sub-module: uses the mechanical sensor at the front end of the puncture needle to collect resistance data during the puncture process in real time, reflecting the interaction between the puncture needle and the surrounding soft tissues;

[0102] The path dynamic adjustment sub-module: integrates the mechanical sensor data into the local reward function, and automatically adjusts the forward direction angle and propulsion force of the puncture needle according to the feedback to achieve dynamic correction of the puncture path;

[0103] An abnormal alarm module is used to monitor abnormal situations such as over-limit resistance, path deviation, and sensor failure during the puncture process in real time; when an abnormality occurs, an alarm is issued through sound, light, pop-up windows, etc., the puncture operation is automatically paused or a safety protection action is executed, and treatment suggestions are provided for medical staff to ensure the safety of the operation;

[0104] The above has shown and described the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification is only the principle of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A soft tissue puncture path planning and monitoring method based on electrical impedance imaging, characterized by: Step S1: Before puncture, obtain soft tissue information in different frequency domains through multi-frequency domain electrical impedance imaging, and obtain structural information through ultrasound imaging and magnetic resonance imaging; Step S2, establishing an image reconstruction model, fusing multi-frequency domain electrical impedance imaging information, ultrasound imaging information and nuclear magnetic resonance imaging information, training a priori auxiliary network, and improving resolution with the help of supplementary information; Step S3, preoperative path planning modeling. The preoperative path planning model is based on the reinforcement learning training model. The needle tip position during the puncture process and its relative position to the target point are used as state information. A suitable reward function is defined. A positive reward is given when the needle tip approaches the target point or reaches it smoothly, and a negative reward is given when it collides with an obstacle or deviates from the path. Through continuous interaction with the environment and trial-and-error learning, the intelligent agent can gradually master the optimal action strategy under different states. Step S4: Install a mechanical sensor at the front end of the puncture needle to monitor the resistance of the puncture path in real time during the operation, analyze the state of the surrounding soft tissue through the needle tip sensor, and dynamically adjust the path; The needle tip sensor data N is introduced to optimize the local reward function for real-time monitoring of intraoperative path rationality and abnormality processing; Step S5, iteratively train step S3-step S4 until the model converges, that is, the optimal route design is achieved.

2. The method for soft tissue puncture path planning and monitoring based on electrical impedance imaging according to claim 1, characterized in that: the step S1 further comprises the following contents: Step S11, injecting multi-frequency alternating current through the body surface electrode array to obtain the conductivity distribution in different frequency domains for distinguishing different tissues, and outputting the image in the corresponding frequency domain, recorded as E_f, where f is the frequency; Step S12, obtaining ultrasound imaging data for supplementing tissue structure information before surgery, denoted as U; Step S13: Acquire magnetic resonance imaging data for supplementing high-resolution structural function information before surgery, denoted as M.

3. The method for soft tissue puncture path planning and monitoring based on electrical impedance imaging according to claim 1, characterized in that: The step S2 further includes the following contents: Step S21, combining ultrasound data and nuclear magnetic resonance data to train a priori auxiliary network, extracting tissue features of blood vessels and key anatomical structures of tumors as prior constraints for electrical impedance imaging reconstruction; the priori auxiliary network uses an encoder architecture, with ultrasound imaging data U and nuclear magnetic resonance imaging data M as input, and outputs predicted features of key parts, recorded as a probability map matrix MP, ranging from [0,1], where a tight 0 represents background, and a tight 1 represents blood vessels or key tumor structures, and the calculation formula is as follows: MP=sigmoid(NETWORK(M,U)) Among them, NETWORK represents the encoder layer, sigmoid represents the activation function, and after the ultrasound imaging data U and the magnetic resonance imaging data M pass through the encoder layer and are activated, the prediction probability map MP of the key parts is obtained; The purpose of modeling the loss function is to make the predicted value MP distribution close to the true value MP'. The formula is as follows: Among them, MP represents the feature data predicted by modeling, MP' represents the real feature label, which has only 0 and 1 values, · represents the dot product operation, || represents the sum operation, and μ represents the balance parameter; Step S22, guide the EIT reconstruction network optimization, intelligently fuse the multi-frequency domain electrical impedance imaging information E_f and the prior key part feature matrix MP, use the prior probability map and multi-frequency domain information to improve the resolution and enhance the recognition ability of microstructures; considering the sparsity of the data, the fusion formula is designed as follows: E=MEAN(E_f) Among them, E represents the electrical impedance imaging data after fusion of multiple frequency domains, and MEAN represents the mean operation; Q=Wq·E Where Q represents the mapped electrical impedance imaging data, and Wq is the weight matrix; K=(Wk·MP) T Among them, K represents the mapped prior data MP, which is used to calculate the correlation. For the convenience of calculation, it is transformed using the T transpose operation, and Wk is the corresponding weight matrix; V=Wv·E_f Among them, V is used to calculate the weighted output of correlation, and Wv is the corresponding weight matrix; Among them, A represents the correlation sparse matrix, i, j represent the corresponding positions in the matrix, exp represents the exponential operation, l represents taking only the top-k elements, k is a manually set value, and d represents the scaling dimension; the correlation sparse matrix can focus more on key areas; E'=E+softmax(A)·V Among them, E' represents the output optimized electrical impedance imaging feature, softmax is a normalization function, and the original information E is added to the feature map softmax(A)·V after calculating the correlation to obtain the optimized electrical impedance imaging feature E'.

4. The method for soft tissue puncture path planning and monitoring based on electrical impedance imaging according to claim 1, characterized in that: The step S3 further includes the following contents: Step S31: Based on the optimized electrical impedance imaging feature E, the key areas are detected and marked by threshold segmentation and morphological operation, and the key areas are divided into high-risk areas and target tissues; Step S32: Path planning modeling Modeling goal: complete optimal path planning Define the action space A: {forward direction angle, propulsion force}, Define the state space S: {impedance imaging feature E', needle tip position, high-risk area and target tissue} Global Reward Function Design Globalr = R target proximity + R danger zone penalty + R age coefficient The global reward function is used to guide the puncture path to avoid macroscopic obstacles and approach the target efficiently. The target proximity encourages rapid approach to the target tissue, and the dangerous area penalty punishes the path for approaching high-risk areas to ensure a safe distance. The age factor adjusts the strategy according to the patient's age; R target proximity = -a ||p_cur-p_obj||2 R target proximity encourages rapid approach to the target tissue and is calculated as the Euclidean distance from the needle tip to the target tissue, where p_cur represents the needle tip position, p_obj represents the target position, ||||2 represents the Euclidean distance, and a is the item weight; R Dangerous area penalty = -ρ∑ i∈高危区域 ||p_cur-p_danger_i||2 R Dangerous area penalty punishes paths that are close to high-risk areas, i represents the high-risk area number, p_danger_i represents the location of the ith high-risk area, ||||2 represents the Euclidean distance, and ρ represents the penalty coefficient; R Age coefficient = -c·|age-30| The R age coefficient adjusts the promotion strategy according to the patient's age. age represents the patient's age, c represents the age item weight, and || is the absolute value. When the patient is an elderly person or a child, the promotion intensity needs to be reduced. Local reward function design; Localr = R path smoothness The local reward function is used to refine the design and balance path efficiency and safety. The R path smoothness encourages smooth changes in the puncture direction of adjacent actions to reduce the risk of tissue tearing caused by sharp turns. R path smoothness = -τ∑ i∈动作序列 |θcur_i-θpre_i| R path smoothness encourages the smooth change of puncture direction in adjacent action sequences. i represents a certain action, θcur_i represents the forward direction angle at the i-th time, θcur_i represents the previous forward direction angle at the i-th time, || belongs to the absolute value operation, and τ represents the weight of smoothness; The reward function is the sum of the global and local reward functions; R=Globalr+Localr The Value function update strategy is as follows: V(s t ,a t )←V(s t ,a t )+[R+σmax a' V(s t+1 ,a')-V(s t ,a t )] The Value function V is the key to modeling learning, t represents the time t, s t represents the state at time t, a t represents the action taken at time t, R is the above reward function, σ is the discount factor, max a' V(s t+1 ,a') represents selecting an action a', and the next state after executing the action is s t+1 The corresponding Value function is the largest.

5. The method for soft tissue puncture path planning and monitoring based on electrical impedance imaging according to claim 1, characterized in that: said step S4 further comprises the following contents: The local reward function is optimized as: Localr=Localr+R_resistance abnormality R_resistance abnormality = -d·(Ft-Fsafe) It measures whether the resistance Ft at the current moment t exceeds the expected safety threshold Fsafe, and imposes a penalty according to the excess ratio. d is the penalty weight of the resistance anomaly item.

6. A soft tissue puncture path planning and monitoring system based on electrical impedance tomography, the system being used to execute the soft tissue puncture path planning and monitoring method based on electrical impedance tomography according to any one of claims 1 to 4, characterized in that: The data acquisition module is used to inject multi-frequency alternating current through the body surface electrode array to obtain conductivity distribution data in different frequency domains; it also collects ultrasound imaging data and magnetic resonance imaging data to provide a multi-modal information basis for subsequent path planning, and ensures the temporal and spatial consistency and availability of each modality data through data preprocessing; The image reconstruction module is used to construct a priori auxiliary network, extract key anatomical structure features of blood vessels and tumors based on ultrasound and MRI data, and generate a probability map matrix; at the same time, it guides the EIT reconstruction network, fuses multi-frequency domain electrical impedance imaging information with the prior feature matrix, and uses weighted and correlation calculation strategies to improve the resolution and microstructure recognition capabilities of electrical impedance imaging; A route planning module is used to mark high-risk areas and target tissues through threshold segmentation and morphological operations based on image reconstruction results; Construct a reinforcement learning model to define the state space including impedance characteristics and needle tip position elements and the action space including the forward direction angle and propulsion force; design global and local reward functions, combine patient age, target proximity, and distance to dangerous areas, and iterate and optimize the Value function to plan a safe and efficient puncture path; The intraoperative monitoring module is used to collect resistance, resistance change rate and lateral force data in real time using the mechanical sensor at the front end of the puncture needle; Incorporate sensor data into the local reward function to dynamically adjust the puncture path; Analyze data trends, predict potential risks, and achieve real-time dynamic monitoring and path correction during the puncture process; The abnormal alarm module is used to monitor abnormal conditions such as excessive resistance, path deviation, and sensor failure during the puncture process in real time. When an abnormality occurs, an alarm is issued through sound, light, or pop-up windows, automatically pausing the puncture operation or executing safety protection actions, and providing treatment suggestions to medical staff to ensure surgical safety.

7. A non-volatile storage medium, characterized in that: The non-volatile storage medium includes a stored program, wherein when the program is executed, the device where the non-volatile storage medium is located is controlled to execute the method according to claim 1.

8. An electronic device, characterized in that: It comprises a processor and a memory; the memory stores computer-readable instructions, and the processor is used to execute the computer-readable instructions, wherein the computer-readable instructions execute the method according to claim 1 when executed.

Images