A robotic automated patch application method and system for peripheral nerve conduction detection

Abstract

This invention relates to the fields of artificial intelligence and biomedical technology, specifically to a robotic automated patching method and system for peripheral nerve conduction detection, as follows: After the doctor selects the examination items on the electromyography instrument, the robot receives instructions and executes the operation; it extracts the patient's limb contour image through a binocular industrial camera system, and calls the nerve-electrode position mapping database to determine the relative positions of standard anatomical feature points; it uses probabilistic Hough transform to filter the lines corresponding to the joints, calculates the intersection points to obtain the feature point image coordinates, combines the relative positions to calculate the target patch position, and then solves the transformation matrix to obtain the three-dimensional coordinates of the robotic arm end; after determining that the electrode inventory is sufficient, it plans a path to grasp the electrodes, aligns and applies them with pressure, verifies the adhesion and contact resistance, and makes fine adjustments if they are unqualified or the impedance drifts; after completing the patching at multiple positioning points, it records the status to form a log, ensuring electrode stability before testing, achieving precise, stable and intelligent patching, reducing medical care intensity and improving diagnostic efficiency.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence and biomedical technology, and in particular to a robotic automated patching method and system for peripheral nerve conduction detection. Background Technology

[0002] Peripheral nervous system diseases are a serious global public health problem affecting human health. According to the World Health Organization, approximately 200 million people worldwide suffer from peripheral neuropathy, with a misdiagnosis rate exceeding 50% for certain specific peripheral neuropathy conditions, posing a significant threat to public health. Peripheral nerve conduction testing requires attaching Ag / AgCl electromyographic electrodes to specific locations on the patient's body surface to collect nerve electrical signals and assess nerve function. However, current nerve conduction testing methods require doctors to perform repeated manual electrode placement based on muscle localization, resulting in subjective biases in localization accuracy. The procedures are time-consuming and labor-intensive, making it difficult to meet the needs of large-scale clinical diagnosis and treatment. Furthermore, this hinders the full automation and intelligentization of peripheral nervous system disease testing, impeding the widespread adoption of high-quality testing technologies in primary healthcare institutions.

[0003] Therefore, developing a fully automated electrode application method and system that can achieve precise positioning, stable attachment, and real-time verification is key to solving the above problems. This invention addresses the issues of high safety requirements for human-computer interaction, poor stability of human-electrode contact, and susceptibility to interference in nerve signal acquisition quality during electrode application. It proposes a robotic automated electrode application method and system for peripheral nerve conduction detection to solve these problems. Summary of the Invention

[0004] This invention addresses the shortcomings of existing technologies by developing a robotic automated patching method and system for peripheral nerve conduction detection. This invention can solve the problems of large positioning deviations, time-consuming and labor-intensive processes, and poor contact stability associated with traditional manual patching for peripheral nerve conduction detection, thereby reducing the workload of medical staff and improving diagnostic efficiency.

[0005] On one hand, the technical solution of this invention to solve the technical problem is a robotic automatic patching method for peripheral nerve conduction detection, comprising the following steps: S1. Extract the patient's limb contour image through a binocular industrial camera system, and then call the neural-electrode position mapping database to determine the relative positional relationship of standard anatomical feature points; S2. Filter out the straight lines corresponding to the joints through probability Hough transform and calculate the intersection of the lines to obtain the coordinates of the feature points in the image coordinate system. Combine the relative positional relationship of the standard anatomical feature points to calculate the position of the target patch. S3. Solve for the transformation matrix between the robot's manipulator base coordinate system and the patient's body surface coordinate system to obtain the three-dimensional coordinates of the robot's end effector corresponding to the target patch position. S4. Determine the inventory of disposable electrodes. If sufficient, use a linear interpolation algorithm to automatically plan the motion path of the robotic arm to grasp the electrodes. S5. Align the electrode and target patch positions. The robotic arm drives the end effector of the gripping electrode to move above the target patch position. S6. The end effector attaches the electrode to the patient's skin and presses it down, and verifies the adhesion; then it detects the skin-electrode contact resistance to determine whether the attachment is qualified. S7. After completing the multi-positioning point patching, an operation log is automatically generated, and the impedance of each positioning point is monitored again. If impedance drift occurs, fine-tuning is performed to ensure that the state of all electrodes is stable before the patient's test begins.

[0006] The process of creating a limb contour image is as follows: A binocular industrial camera system was used to extract limb contours, and a ring of LED lights was deployed around the patient. The color images acquired by the binocular camera were converted to grayscale and smoothed by Gaussian filtering to suppress noise. An 8-neighborhood connection algorithm was used to remove isolated edge points from the extracted edge pixels. By screening the contour area and matching the shape, the complete contours of the patient's upper and lower limbs were obtained, forming a limb contour image.

[0007] The process of determining the target patch position is as follows: For the extracted limb contour image, a probabilistic Hough transform is used to filter out multiple candidate line segments corresponding to joint anatomical structures or bone orientations. Specifically, a list of line segments with consistent length and direction and spatial continuity is selected from the limb contour image. Combined with the prior anatomical information of the detected area, line segment groups that are consistent with the bone axis or the surface contour orientation near the joint are matched from the list of line segments. The equation of the line is determined by fitting. Then, by calculating the intersection point between the lines, the two-dimensional coordinate values ​​of the joint anatomical feature points in the image coordinate system are obtained. The two-dimensional coordinates of the joint anatomical feature points are matched with the relative positions of the pre-stored standard anatomical feature points. Based on the relative distance and orientation between the two points, the offset of the target patch position relative to the joint anatomical feature points is determined, and the two-dimensional coordinates of the target patch position in the image coordinate system are calculated accordingly. Based on the calibration parameters of the binocular camera, the two-dimensional coordinates of the target patch position in the image coordinate system are converted into three-dimensional coordinates in the patient's body surface coordinate system. By solving the homogeneous coordinate transformation matrix between the robot arm base coordinate system and the patient's body surface coordinate system, the target patch position is converted into the corresponding three-dimensional spatial coordinates of the robot arm end effector, which is used to guide the robot arm end effector to reach the target patch position.

[0008] The electrode gripping process is as follows: First, determine if there are available disposable electrodes in the target warehouse. If the inventory is insufficient, trigger a replenishment prompt and suspend the operation. If the inventory is sufficient, use a linear interpolation algorithm to automatically plan the motion path of the end effector and use a 3D depth camera at the end of the robotic arm to scan the surrounding environment of the path in real time to avoid temporary obstacles in time. The robotic arm drives the end effector to move to the target slot position, and uses a magnetic electrode holder to achieve non-destructive gripping of the electrode; During the movement of the robotic arm, warning lights remind medical staff to maintain a reasonable distance.

[0009] The electrode-target patch alignment operation is as follows: (1) In the coarse adjustment phase, the robotic arm drives the end effector of the gripping electrode to move above the preset positioning point, and the electrode is aligned with the area corresponding to the target patch position through the coarse adjustment action; Determine if the coarse adjustment is in place. If it is, trigger the coarse adjustment end signal and switch to the attitude fine adjustment stage. If it is not in place, continue coarse adjustment. If it fails to be in place multiple times, trigger the coarse adjustment abnormality warning and prompt medical staff to check the robotic arm's motion accuracy or obstacles. The criterion for determining whether the coarse adjustment is in place is: the plane deviation between the electrode center and the target patch position in the image coordinate system is less than a preset threshold. (2) During the attitude fine-tuning stage, the rotation angle is fine-tuned to correct the attitude deviation of the end effector; Specifically, the 3D depth camera at the end of the robotic arm captures skin images around the positioning point, collects three evenly distributed sampling points around the positioning point, and uses the least squares method to calculate the equation of the skin reference plane to obtain the fine-tuning rotation angle, which is used to make the electrode contact surface parallel to the patient's skin surface. (3) After the attitude fine-tuning is completed, the superimposed image of the positioning point and the electrode is captured by the 3D camera at the end of the robotic arm, and the contour of the positioning point area is extracted by the image segmentation algorithm. The centroid coordinates of the positioning point and the centroid coordinates of the electrode center are calculated by the centroid method, so as to keep the deviation between the electrode center and the target patch within a reasonable range. For patients with anatomical variations, the alignment system features an external operating handle that allows doctors to manually fine-tune the system as needed.

[0010] The electrode attachment process is as follows: The end effector slowly descends along the normal plane of the measured part, and during the process, the six-dimensional force sensor senses the contact force between the electrode and the skin in real time; When the real-time contact force reaches the preset effective threshold, the end effector maintains the pressure and presses for a moment to allow the electrode to fully adhere to the skin. After the pressure is applied, the end effector rises slightly, and the force sensor uses the pulling force to determine the adhesion level between the electrode and the skin. If the tension rises steadily and reaches the set tension, the electromagnetic clamp releases the electrode, completing the initial attachment; If the tension drops sharply before reaching the set value, it is determined that the adhesion is not firm, and the positioning and alignment steps should be repeated.

[0011] The specific procedure for skin-electrode contact impedance testing is as follows: After the application is complete, the skin-electrode contact resistance is measured by an end effector mounted on the robotic arm. Conduct testing; If the impedance is within the preset optimal range for nerve conduction detection Once the attachment is deemed satisfactory, the process automatically proceeds to the next positioning point. Otherwise, fine-tune the pressure of the end effector on the robotic arm within the positioning point range, and re-test the impedance after each adjustment; if the impedance still does not meet the standard after multiple adjustments, trigger a manual intervention warning and wait for medical staff to check. The process of fine-tuning the pressure is as follows: when When, increase pressure; when Reduce pressure at that time; Pressure adjustment amount during a single adjustment The formula for calculating is: , in, This represents the pressure-resistance correlation coefficient measured experimentally.

[0012] On the other hand, the present invention also provides a robotic automated patching system for peripheral nerve conduction detection, including a module for executing processing instructions for each step of a robotic automated patching method for peripheral nerve conduction detection; Control module: Used to coordinate other modules. It receives patch coordinates from the external system and obtains relevant data from other internal modules, integrates the data, and generates job task instructions. Electrode storage module: It adopts a drawer-type card slot design to store disposable electrodes and feed back the inventory status to the control module; Robot body: It adopts a multi-degree-of-freedom robotic arm structure, receives commands from the control module, and drives the end effector to move from the electrode storage module to the positioning point; the robotic arm body is made of lightweight aluminum alloy, and the surface coating is a medical-grade antibacterial material; Force-tactile feedback end effector: including impedance detection contacts, a six-dimensional force sensor and an electromagnetic electrode holder, for electrode gripping, force-controlled attachment and impedance signal acquisition; the six-dimensional force sensor is used to sense micro-force changes in real time during the attachment process, and the electromagnetic electrode holder is adapted to disposable Ag / AgCl electromyographic electrodes, which perform non-destructive gripping and release by generating electromagnetic attraction, forming a collaborative closed loop with the six-dimensional force sensor and control module; In-situ verification module: integrates impedance measurement circuit, supports multi-channel parallel operation, and is compatible with standard neural conduction instrument interface; conduction is achieved through the impedance detection contact of the force tactile feedback end effector and the attached electrode pad, a test current is applied at the positioning point to detect the skin-electrode contact impedance, and information is linked with the control module until the impedance reaches the standard or an alarm is triggered.

[0013] The effects described in the invention are merely those of the embodiments, and not all the effects of the invention. The above technical solutions have the following advantages or beneficial effects: Existing technologies rely on doctors' experience for traditional manual electrode placement, which can lead to positioning errors and distortion of nerve signal acquisition. This invention uses a combination of "binocular camera + standard anatomical database + coordinate transformation" technology. Through image preprocessing, linear feature screening and precise coordinate calculation, combined with the high-precision control of the robotic arm, it ensures that the electrodes are accurately attached to the detection target area, improves the accuracy of nerve signal acquisition, effectively solves the core problems of strong subjectivity and insufficient accuracy in traditional technology, and reduces the risk of missed diagnosis.

[0014] Existing manual electrode application requires doctors to repeatedly pick up, place, and adjust electrodes, resulting in long application times for each patient and potential electrode contamination or damage during manual transport. This invention, through a design combining a multi-degree-of-freedom robotic arm, a magnetic gripper, and force-controlled application, automates the entire process of electrode grasping, path planning, obstacle avoidance transport, and electrode application. This significantly reduces the repetitive workload of medical staff and addresses the pain points of time-consuming manual operations and easy electrode damage.

[0015] Traditional manual electrode application lacks standardized pressure; insufficient pressure leads to poor electrode contact, while excessive pressure can cause patient discomfort. This invention's end effector integrates a high-resolution six-dimensional force sensor, achieving precise force control during electrode application through closed-loop pressure control. Furthermore, tensile testing verifies adhesion, preventing repetitive work due to electrode detachment during testing, reducing patient discomfort caused by improper pressure, and significantly improving the comfort and continuity of clinical testing.

[0016] In summary, this invention discloses a robotic automated patching method and system for peripheral nerve conduction detection, aiming to solve the technical pain points of traditional manual patching, such as low positioning accuracy, long operation time, and reliance on doctor experience. This invention uses a standard anatomical database as its core guide, employing machine vision technology to achieve personalized conversion from standard anatomical target detection locations to the patient's actual anatomical points, accurately matching the anatomical characteristics of different patients. Relying on closed-loop control logic, it automatically and systematically completes the entire process of "attachment positioning point detection, electrode patch grasping, electrode patch alignment with attachment points, force-controlled electrode patch attachment, impedance verification and fine-tuning, and multi-point cyclic attachment," without repeated manual intervention. This invention reuses existing multi-degree-of-freedom robotic arms to complete core motion functions, combined with a six-dimensional force sensor to achieve precise force control during the attachment process, ensuring stable electrode-skin contact and controllable comfort. The hardware and software adopt a highly modular design, with positioning, grasping, attachment, and verification modules working independently and collaboratively, reducing overall design complexity and maintenance costs. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0018] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0019] Figure 2 This is a schematic diagram of a clinical operation scenario for the robotic automated patch application system proposed in this invention. Detailed Implementation

[0020] To clearly illustrate the technical features of this solution, the invention will be described in detail below through specific embodiments and in conjunction with the accompanying drawings. The following disclosure provides many different embodiments or examples for implementing different structures of the invention. To simplify the disclosure of the invention, the components and arrangements of specific examples are described below.

[0021] Example 1 A robotic automated patching method for peripheral nerve conduction detection involves a doctor selecting the tests to be performed on an electromyography (EMG) instrument based on the patient's condition, and the robot receiving the test instructions. The EMG instrument supports tests including motor nerve conduction velocity detection, sensory nerve conduction velocity detection, F-wave detection, and H-reflex detection. Motor / sensory nerve conduction detection requires the attachment of 2-3 electrodes, including a proximal stimulation electrode, a distal stimulation electrode, a recording electrode, and a reference electrode. F-wave / H-reflex detection also requires the attachment of 2-3 electrodes. Upper limb detection focuses on the median nerve and ulnar nerve, while lower limb detection focuses on the common peroneal nerve, tibial nerve, and sural nerve.

[0022] like Figure 1As shown, the specific operation steps of a robotic automated patching method for peripheral nerve conduction detection are as follows: S1. Extract the patient's limb contour image through a binocular industrial camera system, and then call the neural-electrode position mapping database to determine the relative positional relationship of standard anatomical feature points; S2. Filter out the straight lines corresponding to the joints through probability Hough transform and calculate the intersection of the lines to obtain the coordinates of the feature points in the image coordinate system. Combine the relative positional relationship of the standard anatomical feature points to calculate the position of the target patch. S3. Solve for the transformation matrix between the robot's manipulator base coordinate system and the patient's body surface coordinate system to obtain the three-dimensional coordinates of the robot's end effector corresponding to the target patch position. S4. Determine the inventory of disposable electrodes. If sufficient, use a linear interpolation algorithm to automatically plan the motion path of the robotic arm to grasp the electrodes. S5. Align the electrode and target patch positions. The robotic arm drives the end effector of the gripping electrode to move above the target patch position. S6. The end effector attaches the electrode to the patient's skin and presses it down, and verifies the adhesion; then it detects the skin-electrode contact resistance to determine whether the attachment is qualified. S7. After completing the multi-positioning point patching, an operation log is automatically generated. The operation log records information such as the positioning point number, patching time, patching pressure, impedance value, and qualification status, and is stored in CSV file format. The impedance of each positioning point is monitored again. If impedance drift occurs, fine-tuning is performed to ensure that the state of all electrodes is stable before the patient's test begins.

[0023] In a specific implementation, the limb contour image process is as follows: A binocular industrial camera system was used to extract the limb contour. The binocular industrial camera was deployed at a height of 30-50cm from the patient's limb. The acquisition frame rate was set, and a ring of LED fill lights was deployed around the patient. The fill light intensity was dynamically adjusted according to the patient's skin color. The color images captured by the binocular camera are converted to grayscale, and the images are smoothed and noise is suppressed by Gaussian filtering; An 8-neighborhood connection algorithm is used to remove isolated edge points from the extracted edge pixels. By filtering the contour area and matching the shape, the complete contours of the patient's upper and lower limbs are obtained, forming a limb contour image.

[0024] In a specific implementation, the process of determining the target patch position is as follows: For the extracted limb contour images, a probabilistic Hough transform is used to filter out multiple candidate line segments corresponding to joint anatomy or bone orientation. Specifically, a list of line segments with consistent length, direction, and spatial continuity is selected from the limb contour images. Combined with prior anatomical information of the detected area, line segment groups that are consistent with the bone axis or the surface contour of the body near the joint are matched from the list of line segments. The equation of the corresponding line is determined by fitting. Then, by calculating the intersection points between the lines, the two-dimensional coordinate values ​​of the joint anatomical feature points in the image coordinate system are obtained. Among them, the angle range of the line corresponding to the elbow joint is 120°-150°, the length of the line corresponding to the knee joint accounts for 30%-40% of the total length of the limb contour, and the angle of the line corresponding to the shoulder joint is 45°-90°. The two-dimensional coordinates of the joint anatomical feature points are matched with the relative positions of the pre-stored standard anatomical feature points. Based on the relative distance and orientation between the two points, the offset of the target patch position relative to the joint anatomical feature points is determined, and the two-dimensional coordinates of the target patch position in the image coordinate system are calculated accordingly. Based on the calibration parameters of the binocular camera, the two-dimensional coordinates of the target patch position in the image coordinate system are converted into three-dimensional coordinates in the patient's body surface coordinate system. By solving the homogeneous coordinate transformation matrix between the robot arm base coordinate system and the patient's body surface coordinate system, the target patch position is converted into the corresponding three-dimensional spatial coordinates of the robot arm end effector, which is used to guide the robot arm end effector to reach the target patch position.

[0025] In a specific implementation, the electrode gripping process is as follows: First, determine if there are available disposable electrodes in the target warehouse. If the inventory is insufficient, trigger a replenishment prompt and suspend the operation. If the inventory is sufficient, use a linear interpolation algorithm to automatically plan the motion path of the end effector and use a 3D depth camera at the end of the robotic arm to scan the surrounding environment of the path in real time to avoid temporary obstacles in time. The robotic arm drives the end effector to move to the target slot position and uses a magnetic electrode holder to grasp the electrode without damage. The end effector is equipped with a magnetic electrode holder, which is the core execution component of the end effector and is used to grasp the electrode. The end effector is also equipped with impedance detection contacts and pressure sensors for impedance detection and contact force acquisition. In addition, warning lights remind medical staff to maintain a reasonable distance during the movement of the robotic arm.

[0026] In a specific implementation, the electrode-target patch alignment operation is as follows: (1) In the coarse adjustment phase, the robotic arm drives the end effector of the gripping electrode to move above the preset positioning point, and the electrode is aligned with the area corresponding to the target patch position through the coarse adjustment action; Determine if the coarse adjustment is in place. If it is, trigger the coarse adjustment end signal and switch to the attitude fine adjustment stage. If it is not in place, continue coarse adjustment. If it fails to be in place multiple times, trigger the coarse adjustment abnormality warning and prompt medical staff to check the robotic arm's motion accuracy or obstacles. The criterion for determining whether the coarse adjustment is in place is: the plane deviation between the electrode center and the target patch position in the image coordinate system is less than a preset threshold. (2) During the attitude fine-tuning stage, the rotation angle is fine-tuned to correct the attitude deviation of the end effector; Specifically, the 3D depth camera at the end of the robotic arm captures skin images around the positioning point, collects three evenly distributed sampling points around the positioning point, and uses the least squares method to calculate the equation of the skin reference plane to obtain the fine-tuning rotation angle, which is used to make the electrode contact surface parallel to the patient's skin surface. (3) After the attitude fine-tuning is completed, the superimposed image of the positioning point and the electrode is captured by the 3D camera at the end of the robotic arm, and the contour of the positioning point area is extracted by the image segmentation algorithm. The centroid coordinates of the positioning point and the centroid coordinates of the electrode center are calculated by the centroid method, so as to keep the deviation between the electrode center and the target patch within a reasonable range. For patients with anatomical variations, the alignment system features an external operating handle that allows doctors to manually fine-tune the system as needed.

[0027] In a specific implementation, the electrode attachment operation is as follows: The end effector slowly descends along the normal plane of the measured part, and during the process, the six-dimensional force sensor senses the contact force between the electrode and the skin in real time; When the real-time contact force reaches the preset effective threshold, the end effector maintains the pressure and presses for a moment to allow the electrode to fully adhere to the skin. After the pressure is applied, the end effector rises slightly, and the force sensor uses the pulling force to determine the adhesion level between the electrode and the skin. If the tension rises steadily and reaches the set tension, the electromagnetic clamp releases the electrode, completing the initial attachment; If the tension drops sharply before reaching the set value, it is determined that the adhesion is not firm, and the positioning and alignment steps should be repeated.

[0028] In a specific implementation, the skin-electrode contact impedance detection operation is as follows: After the application is complete, the skin-electrode contact resistance is measured by an end effector mounted on the robotic arm. Conduct testing; If the impedance is within the preset optimal range for nerve conduction detection Once the attachment is deemed satisfactory, the process automatically proceeds to the next positioning point. Otherwise, fine-tune the pressure of the end effector on the robotic arm within the positioning point range, and re-check the impedance after each adjustment; if the impedance still does not meet the standard after multiple adjustments, trigger a manual intervention warning and wait for medical personnel to examine the patient. The process of fine-tuning the pressure is as follows: when When, increase pressure; when Reduce pressure at that time; Pressure adjustment amount during a single adjustment The formula for calculating is: , in, This represents the pressure-resistance correlation coefficient measured experimentally.

[0029] In specific applications, the following precautions should be taken when using the method of this invention: Before starting the device, check the cleanliness of the lenses of the binocular camera and the 3D depth camera, and the ring LED. The supplemental lighting is undamaged, the robotic arm joints move flexibly without jamming, and the electrode compartment is free of foreign objects. Patients must remove skin oil and hair from the patch area, avoid wearing metal jewelry, and position their limbs naturally and relaxed to avoid obscuring joint features. When grasping electrodes, ensure that disposable electrodes are not expired or damaged, and that the magnetic gripper adheres to the non-conductive edge of the electrode to prevent damage to the conductive layer. During robotic arm movement, medical staff must observe from a safe distance throughout, and are prohibited from touching the moving robotic arm or electrodes to avoid collisions with the patient's limbs. When manually fine-tuning, press the control handle buttons gently to avoid rapid, large adjustments. After fine-tuning for patients with anatomical variations, impedance must be retested to ensure compliance with standards. If impedance testing repeatedly fails to meet standards, check for electrode damage, dry skin, or dirt. If necessary, replace the electrode and clean the skin before reapplying. After the procedure, clean the electrode compartment promptly, replenish disposable electrodes, and export the work log before turning off the equipment to prevent data loss. The equipment requires regular calibration: the binocular camera, robotic arm positioning accuracy, and impedance detection module all require regular calibration, and calibration records must be archived for future reference.

[0030] Example 2 A robotic automated patching system for peripheral nerve conduction detection includes modules for executing processing instructions for each step of a robotic automated patching method for peripheral nerve conduction detection. The system includes the following modules: The control module coordinates with other modules, receiving the patch coordinates from the external system and data such as the patient's upper and lower limb contours, robotic arm pose, contact force, impedance, and inventory from other internal modules to complete tasks such as path planning, pressure control, and fine-tuning command output, thereby achieving full automation of the electrode gripping, attachment, and verification process. The electrode storage module adopts a drawer-type slot design to store disposable electrodes and feed back the inventory status to the control module; The robot body adopts a multi-degree-of-freedom robotic arm structure, receives instructions from the control module, and drives the end effector to move from the electrode storage module to the positioning point; the robotic arm should support operation in confined clinical environments, and the arm body material is lightweight aluminum alloy with a surface coating of medical-grade antibacterial material; The force-haptic feedback end effector includes a six-dimensional force sensor, an electromagnetic electrode holder, and impedance detection contacts, enabling electrode gripping, force-controlled attachment, and impedance signal acquisition. The integrated six-dimensional force sensor is installed between the electrode holder base and the robotic arm end effector to detect minute force changes during attachment in real time and ensures the accuracy and safety of the attachment action based on closed-loop control. The electromagnetic electrode holder is the core execution component of the end effector, compatible with disposable Ag / AgCl electromyographic electrodes. It generates electromagnetic attraction to achieve non-destructive gripping and release, forming a collaborative closed loop with the six-dimensional force sensor and control module. The in-situ verification module integrates an impedance measurement circuit, supports multi-channel parallel operation, and is compatible with standard neural conduction instrument interfaces. It achieves conductivity between the impedance detection contacts of the end effector and the attached electrode pads, applies a test current at the positioning point, detects the skin-electrode contact impedance, and links with the control module until the impedance meets the standard or an alarm is triggered.

[0031] Example 3 like Figure 2 The diagram shows a clinical scenario of a robotic automatic patch application system. Medical staff need to guide the patient to lie on the testing bed and adjust them to a comfortable and fixed position. After the position is confirmed, the doctor operates the electromyography (EMG) machine to select the test items. The EMG machine is connected to the robot, which is equipped with a robotic automatic patch application system. The robot is controlled to apply patches to the patient according to the instructions of the EMG machine.

[0032] In the specific implementation process, taking ulnar nerve conduction detection as an example, the robot body can be a commercially available collaborative robot with more than 6 degrees of freedom, with a repeatability error ≤ ±0.1mm; the force sensor module can be a high-resolution six-dimensional force sensor, such as the ATI Mini / Nano series, etc.; the disposable electrodes can be specific Ag / AgCl electromyographic electrodes, which can achieve electromagnetic adsorption. In this case, the patient lies supine on the testing table with the right arm extended (palm facing up), exposing the hypothenar eminence, wrist, and elbow areas; after the operator wipes the above areas with alcohol, the automatic patching system is started, and the "ulnar nerve detection" mode is selected in the host computer software of the control module (this detection task defaults to attaching 2 "recording electrodes" and 1 "ground electrode"). The specific implementation steps of this invention are as follows: (1) The control module starts a self-test to determine whether the power supply and communication of each module are normal; through communication with the external system or input through the human-machine interface, the three-dimensional coordinates and rotation angle of the end of the robotic arm for the hypothenar muscle, wrist stimulation point and elbow stimulation point of the ulnar nerve detection are determined.

[0033] (2) The control module plans the motion path of the end effector based on the inventory signal of the electrode storage module; drives the end effector to move to the target slot and grabs a disposable electrode through the electromagnetic electrode holder.

[0034] (3) The robotic arm drives the end effector to move directly above the positioning coordinates, and the end effector adjusts its attitude to ensure that the deviation between the electrode center and the positioning point is within a reasonable range.

[0035] (4) Electrode force control attachment: The end effector slowly descends, and the six-dimensional force sensor provides real-time feedback on the contact force; when the contact pressure reaches the set threshold, it is pressed steadily for a moment; thereafter, the end effector rises slightly to determine whether the electrode adhesion meets the standard; after meeting the standard, the electrode holder releases the electrode to complete the initial attachment.

[0036] (5) The impedance monitoring module detects the electrode-skin contact impedance at the positioning point. If the impedance is within the optimal range for nerve conduction detection, the patch is deemed to be qualified and the next positioning point patching process is initiated. If the impedance is not qualified, the control module drives the end effector to increase or decrease the pressure within the positioning point range until the impedance meets the standard, or issues an early warning and prompts for manual intervention.

[0037] (6) Repeat steps (2)-(5) to complete the attachment of the “recording electrode” and “ground electrode” required for this test at the corresponding positioning points.

[0038] (7) The impedance monitoring module continuously monitors for 2 minutes. If impedance drift occurs, the control module sends a fine-tuning command in real time to ensure that the state of all electrodes is stable before the test begins.

[0039] Once all electrode impedances have stabilized, a "ready" signal is sent to the nerve conduction detector, awaiting the staff to connect the attached electrode pads and electromyography instrument to begin the formal patient testing.

[0040] Although the specific embodiments of the invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the invention. Based on the technical solutions of the invention, various modifications or variations that can be made by those skilled in the art without creative effort are still within the scope of protection of the invention.

Claims

1. A robotic automated patching method for peripheral nerve conduction testing, characterized by, After the doctor selects the examination items on the electromyography instrument based on the patient's condition, the robot receives the instructions and performs the following operations: S1. Extract the patient's limb contour image through a binocular industrial camera system, and then call the neural-electrode position mapping database to determine the relative positional relationship of standard anatomical feature points; S2. Filter out the straight lines corresponding to the joints through probability Hough transform and calculate the intersection of the lines to obtain the coordinates of the feature points in the image coordinate system. Combine the relative positional relationship of the standard anatomical feature points to calculate the position of the target patch. S3. Solve for the transformation matrix between the robot's manipulator base coordinate system and the patient's body surface coordinate system to obtain the three-dimensional coordinates of the robot's end effector corresponding to the target patch position. S4. Determine the inventory of disposable electrodes. If sufficient, use a linear interpolation algorithm to automatically plan the motion path of the robotic arm to grasp the electrodes. S5. Align the electrode and target patch positions. The robotic arm drives the end effector of the gripping electrode to move above the target patch position. S6. The end effector attaches the electrode to the patient's skin and presses it down, and verifies the adhesion; then it detects the skin-electrode contact resistance to determine whether the attachment is qualified. S7. After completing the multi-positioning point patching, an operation log is automatically generated, and the impedance of each positioning point is monitored again. If impedance drift occurs, fine-tuning is performed to ensure that the state of all electrodes is stable before the patient's test begins.

2. The robotic automated patching method for peripheral nerve conduction testing of claim 1, wherein, The process of creating a limb contour image is as follows: A binocular industrial camera system was used to extract limb contours, and a ring of LED lights was deployed around the patient. The color images acquired by the binocular camera were converted to grayscale and smoothed by Gaussian filtering to suppress noise. An 8-neighborhood connection algorithm was used to remove isolated edge points from the extracted edge pixels. By screening the contour area and matching the shape, the complete contours of the patient's upper and lower limbs were obtained, forming a limb contour image.

3. The robotic automated patching method for peripheral nerve conduction testing of claim 1, wherein, The process of determining the target patch position is as follows: For the extracted limb contour image, the probability Hough transform is used to filter out multiple candidate line segments corresponding to the joint anatomical structure or bone direction. Specifically, a list of line segments with consistent length and direction and spatial continuity is selected from the limb contour image. Combined with the prior anatomical information of the detected part, the line segment group that is consistent with the bone axis or the surface contour direction near the joint is matched from the list of line segments. The equation of the line is determined by fitting method. Then, by calculating the intersection points between the straight lines, the two-dimensional coordinates of the joint anatomical feature points in the image coordinate system are obtained; The two-dimensional coordinates of the joint anatomical feature points are matched with the relative positions of the pre-stored standard anatomical feature points. Based on the relative distance and orientation between the two points, the offset of the target patch position relative to the joint anatomical feature points is determined, and the two-dimensional coordinates of the target patch position in the image coordinate system are calculated accordingly. Based on the calibration parameters of the binocular camera, the two-dimensional coordinates of the target patch position in the image coordinate system are converted into three-dimensional coordinates in the patient's body surface coordinate system. By solving the homogeneous coordinate transformation matrix between the robot arm base coordinate system and the patient's body surface coordinate system, the target patch position is converted into the corresponding three-dimensional spatial coordinates of the robot arm end effector, which is used to guide the robot arm end effector to reach the target patch position.

4. The robotic automated patching method for peripheral nerve conduction detection according to claim 1, characterized in that, Electrode gripping The process is as follows: First, determine if there are available disposable electrodes in the target warehouse. If the inventory is insufficient, trigger a replenishment prompt and suspend the operation. If the inventory is sufficient, use a linear interpolation algorithm to automatically plan the motion path of the end effector and use a 3D depth camera at the end of the robotic arm to scan the surrounding environment of the path in real time to avoid temporary obstacles in time. The robotic arm drives the end effector to move to the target slot position, and uses a magnetic electrode holder to achieve non-destructive gripping of the electrode; During the movement of the robotic arm, warning lights remind medical staff to maintain a reasonable distance.

5. The robotic automated patching method for peripheral nerve conduction detection according to claim 1, characterized in that, The electrode-target patch alignment operation is as follows: (1) In the coarse adjustment phase, the robotic arm drives the end effector of the gripping electrode to move above the preset positioning point, and the electrode is aligned with the area corresponding to the target patch position through the coarse adjustment action; Determine if the coarse adjustment is complete. If it is, trigger the coarse adjustment end signal and switch to the attitude fine adjustment stage. If it is not in place, continue to make coarse adjustments. If it is not in place multiple times, trigger a coarse adjustment abnormality warning, prompting medical staff to check the robotic arm's motion accuracy or obstacles. The criterion for determining whether the coarse adjustment is in place is: the plane deviation between the electrode center and the target patch position in the image coordinate system is less than a preset threshold. (2) During the attitude fine-tuning stage, the rotation angle is fine-tuned to correct the attitude deviation of the end effector; Specifically, the 3D depth camera at the end of the robotic arm captures skin images around the positioning point, collects three evenly distributed sampling points around the positioning point, and uses the least squares method to calculate the equation of the skin reference plane to obtain the fine-tuning rotation angle, which is used to make the electrode contact surface parallel to the patient's skin surface. (3) After the attitude fine-tuning is completed, the superimposed image of the positioning point and the electrode is captured by the 3D camera at the end of the robotic arm, and the contour of the positioning point area is extracted by the image segmentation algorithm. The centroid coordinates of the positioning point and the centroid coordinates of the electrode center are calculated by the centroid method, so as to keep the deviation between the electrode center and the target patch within a reasonable range. For patients with anatomical variations, the alignment system features an external operating handle that allows doctors to manually fine-tune the system as needed.

6. The robotic automated patching method for peripheral nerve conduction detection according to claim 1, characterized in that, The electrode attachment process is as follows: The end effector slowly descends along the normal plane of the measured part, and during the process, the six-dimensional force sensor senses the contact force between the electrode and the skin in real time; When the real-time contact force reaches the preset effective threshold, the end effector maintains the pressure and presses for a moment to allow the electrode to fully adhere to the skin. After the pressure is applied, the end effector rises slightly, and the force sensor uses the pulling force to determine the adhesion level between the electrode and the skin. If the tension rises steadily and reaches the set tension, the electromagnetic clamp releases the electrode, completing the initial attachment; If the tension drops sharply before reaching the set value, it is determined that the adhesion is not firm, and the positioning and alignment steps should be repeated.

7. The robotic automated patching method for peripheral nerve conduction detection according to claim 1, characterized in that the skin... The specific procedure for electrode contact impedance detection is as follows: After the application is complete, the skin-electrode contact resistance is measured by an end effector mounted on the robotic arm. Conduct testing; If the impedance is within the preset optimal range for nerve conduction detection Once the attachment is deemed satisfactory, the process automatically proceeds to the next positioning point. Otherwise, fine-tune the pressure of the end effector on the robotic arm within the positioning point range, and re-check the impedance after each adjustment; if the impedance still does not meet the standard after multiple adjustments, trigger a manual intervention warning and wait for medical personnel to examine the patient. The process of fine-tuning the pressure is as follows: when When, increase pressure; when Reduce pressure at that time; Pressure adjustment amount during a single adjustment The formula for calculating is: ， in, This represents the pressure-resistance correlation coefficient measured experimentally.

8. A robotic automated patching system for peripheral nerve conduction detection, characterized in that: The system is installed in the robot body to perform the robotic automatic patching method for peripheral nerve conduction detection as described in any one of claims 1 to 7. The system includes the following modules: Control module: Used to coordinate other modules. It receives patch coordinates from the external system and obtains relevant data from other internal modules, integrates the data, and generates job task instructions. Electrode storage module: It adopts a drawer-type card slot design to store disposable electrodes and feed back the inventory status to the control module; Robot body: It adopts a multi-degree-of-freedom robotic arm structure, receives commands from the control module, and drives the end effector to move from the electrode storage module to the positioning point; the robotic arm body is made of lightweight aluminum alloy, and the surface coating is a medical-grade antibacterial material; Force-tactile feedback end effector: including impedance detection contacts, a six-dimensional force sensor and an electromagnetic electrode holder, for electrode gripping, force-controlled attachment and impedance signal acquisition; the six-dimensional force sensor is used to sense micro-force changes in real time during the attachment process, and the electromagnetic electrode holder is adapted to disposable Ag / AgCl electromyographic electrodes, which perform non-destructive gripping and release by generating electromagnetic attraction, forming a collaborative closed loop with the six-dimensional force sensor and control module; In-situ verification module: integrates impedance measurement circuit, supports multi-channel parallel operation, and is compatible with standard neural conduction instrument interface; conduction is achieved through the impedance detection contact of the force tactile feedback end effector and the attached electrode pad, a test current is applied at the positioning point to detect the skin-electrode contact impedance, and information is linked with the control module until the impedance reaches the standard or an alarm is triggered.

Images