Transcranial direct current stimulation electrode positioning method based on artificial intelligence image recognition and application

Abstract

The invention provides a transcranial direct current stimulation electrode positioning method based on artificial intelligence image recognition and application. In order to solve the technical problem that it is difficult for a user to autonomously and accurately place tDCS electrodes in a family environment, the method comprises the steps that real-time face image data of the user is obtained through a camera of intelligent equipment; using an artificial intelligence model to identify and output coordinates of a plurality of key points of the face; calculating the target position of the electrode on the head through an algorithm based on the geometrical relationship of the key points; and generating a visual mark on a display interface, and guiding a user to adjust the electrode to the target position in real time. The method can also comprise the steps of performing electrode attachment detection through a sub-electrode, and performing dual positioning verification by using an Apri lTag two-dimensional code. According to the method, a user can conveniently and quickly complete high-precision electrode positioning, and the usability, the treatment accuracy and the safety of tDCS equipment in a family scene are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical equipment, in particular to an electrode positioning technology of transcranial direct current stimulation (tDCS), and especially to an electrode automatic positioning method based on artificial intelligence image recognition and processing and application. BACKGROUND

[0002] Transcranial direct current stimulation (tDCS) is a non-invasive neuromodulation technique that applies a weak, constant direct current through electrodes placed on the scalp to modulate the excitability of the cerebral cortex. It is widely used in the clinical treatment and research of neuropsychiatric diseases such as depression, anxiety, and chronic pain.

[0003] The therapeutic effect of tDCS is highly dependent on the accuracy of electrode placement. For example, when treating depression, the cathode electrode needs to be accurately placed on the left dorsolateral prefrontal cortex (L-DLPFC), which corresponds to the F3 point in the international 10-20 electroencephalogram (EEG) system. In a clinical environment, the positioning of the electrode is usually completed by medical personnel who have been trained professionally. They use tools such as tape measures, head caps, and marker pens to manually determine the electrode position according to the 10-20 system measurement method. This process is tedious, time-consuming, and highly dependent on the experience of the operator.

[0004] With the development of tDCS devices towards home and personalization, how to enable non-professional users to independently and accurately complete electrode positioning without external guidance has become a technical problem that needs to be solved. Existing home devices usually adopt a fixed headband or headcap design, but due to significant differences in head shape, head circumference, and facial structure among individuals, this "one-size-fits-all" solution cannot guarantee the accuracy of positioning, which may cause the stimulation target to deviate, affecting the treatment effect, and even causing unintended side effects.

[0005] Therefore, developing a technical solution that enables users to independently, conveniently, and accurately complete electrode positioning in a home scenario is crucial for improving the accessibility and effectiveness of tDCS treatment. SUMMARY

[0006] The present application provides a transcranial direct current stimulation electrode positioning method based on artificial intelligence image recognition and application, which addresses the problem that existing tDCS electrode positioning technologies lack individualized, accurate, and easy-to-operate self-positioning methods in home application scenarios.

[0007] The core technology of the present application is mainly to use the camera of the intelligent device to collect the facial image of the user in real time, identify the key anatomical anchor points of the face through the artificial intelligence model, and based on the geometric relationship of these anchor points, calculate the two-dimensional coordinates of the personalized and accurate tDCS electrode target point on the head through the algorithm, and finally guide the user to complete the positioning in a visual way.

[0008] In a first aspect, the present application provides a transcranial direct current stimulation electrode positioning method based on artificial intelligence image recognition, which comprises the following steps: acquiring real-time image data of the user's face; using an artificial intelligence face recognition model to process the real-time image data, identifying and outputting the coordinates of multiple preset key points of the user's face, including the feature points of the user's jawline, eyebrows, eyes, nose, and lip area; Based on the coordinates of the multiple preset key points, at least one target position coordinate of the transcranial direct current stimulation electrode on the user's head is calculated through a preset geometric calculation model; On the display interface, visual guidance information is generated according to the target position coordinates to guide the user to place the transcranial direct current stimulation electrode at the target position.

[0009] Further, the step of calculating the target position coordinate by the geometric calculation model comprises: According to the distance between at least two key points, a head size parameter for representing the size of the user's head is calculated; According to the relative position of the user's two eye pupil key points, a head tilt parameter for representing the tilt state of the head is determined; According to the key points of the eyebrows or nose, a first reference point on the facial midline is determined; Based on the first reference point, other key points of the face, and the head size parameter, a second reference point on the cranial midline is calculated; Combined with the second reference point, the head size parameter and the head tilt parameter, the target position coordinate is calculated.

[0010] Further, the target position is the area corresponding to the left dorsolateral prefrontal cortex or the right dorsolateral prefrontal cortex, which is labeled as F3 or F4 point in the international 10-20 brain electrical system.

[0011] Further, it further comprises a double positioning verification step: Through the machine readable code provided on the shell of the image recognition device, the real-time spatial pose information of the transcranial direct current stimulation device is obtained; Compare and verify the spatial pose information with the target position calculated by the facial key points to confirm that the actual placement position of the transcranial direct current stimulation electrode is consistent with the target position.

[0012] Further, the machine-readable code is an April Tag two-dimensional code.

[0013] Further, the artificial intelligence face recognition model is an OpenPose model based on a convolutional neural network.

[0014] Further, before acquiring the real-time image data of the user's face, at least one of the following electrical detection steps is further included: Electrode patch preliminary detection: by detecting the current parameters between the sub-electrodes inside the electrode, it is judged whether the disposable electrode patch and the transcranial direct current stimulation electrode are attached well; Head overall attachment detection: by applying a preset waveform current between the cathode electrode and the anode electrode and detecting, it is judged whether the transcranial direct current stimulation device as a whole is attached well to the user's head.

[0015] In a second aspect, the present application provides a transcranial direct current stimulation electrode positioning device based on artificial intelligence image recognition, comprising: An electrode assembly comprising a cathode and an anode, the cathode and the anode being provided with sub-electrodes and a conductive medium; A current detection module for detecting the current between the sub-electrodes and the current between the cathode and the anode to evaluate the patch adhesion and the head overall attachment state; An image acquisition module for acquiring real-time image data of the user's face; An artificial intelligence processing module for identifying a preset key point of the user's face in the real-time image data, and calculating at least one target position coordinate of the transcranial direct current stimulation electrode on the user's head based on the preset key point coordinates using a geometric calculation model, the preset key point including the feature points of the user's jawline, eyebrows, eyes, nose and lip area; A visual guidance module for superimposing the target position coordinates in real time to a video output interface to guide the user to place the transcranial direct current stimulation electrode at the target position; A positioning verification module for identifying the positioning mark on the surface of the device, acquiring the pose information of the transcranial direct current stimulation device relative to the head, and performing real-time verification with the face key point information.

[0016] In a third aspect, the present application provides an electronic device comprising a memory and a processor, the memory storing a computer program, and the processor being configured to run the computer program to execute the transcranial direct current stimulation electrode positioning method based on artificial intelligence image recognition described above.

[0017] In a fourth aspect, the present application provides a readable storage medium, the readable storage medium storing a computer program, the computer program comprising program code for controlling a process to execute the process, the process comprising the transcranial direct current stimulation electrode positioning method based on artificial intelligence image recognition described above.

[0018] The main contributions and innovations of the present application are as follows: 1. Breakthrough in professional dependence on family scene: without the guidance of medical staff, users can complete self-positioning through smart phones and wearable tDCS devices, significantly reducing the use threshold and promoting the popularization of tDCS technology at home; 2. Positioning accuracy is significantly improved: relying on the millimeter-level facial key point recognition capability of the OpenPose model, combined with dynamic geometric calculation based on head size and posture, the electrode positioning error is less than 5% of the head diameter, which can accurately match the core target points such as L-DLPFC (F3 point), and ensure the treatment effect; 3. Enhanced safety and reliability: through the dual current feedback mechanism of "patch preliminary detection" and "head overall attachment detection", the poor contact problems such as patch not being attached firmly and hair blocking are excluded in real time, avoiding skin risks caused by local current abnormalities; at the same time, the dual verification of April Tag and facial key points further ensures that the actual electrode position is consistent with the theoretical target point, reducing the probability of invalid stimulation; 4. Strong universality: the technical solution has been verified to be effective in subjects of different countries, races, skin colors and head shapes, and can dynamically adapt to individual facial features and posture differences without the need to adjust hardware or algorithms for specific groups; 5. Convenient and intuitive operation: through the real-time superposition of electrode target point virtual markers on the smart phone video interface, the user can be visually guided to adjust the device position, and the operation process conforms to the usage habits of home users without the need for additional professional training; The details of one or more embodiments of the present application are presented in the following drawings and description, so that other features, objects and advantages of the present application are more concise and easy to understand. BRIEF DESCRIPTION OF DRAWINGS

[0019] The drawings described herein are used to provide further understanding of the present application, and form a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application, and do not constitute an improper limitation on the present application. In the drawings: Figure 1 is a flowchart of a transcranial direct current stimulation electrode positioning method based on artificial intelligence image recognition according to an embodiment of the present application; Figure 2 is an example output diagram of a geometric calculation model according to an embodiment of the present application; Figure 3 is a schematic diagram of a tDCS device according to an embodiment of the present application; Figure 4 is another perspective view of Figure 3 Figure 5 is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present application. ​

[0020] In the figure, 1, the shell body; 2, electrode shell; 3, conductive rubber; 4, headband; 5, switch; 6, LED indicator; 7, charging port. DETAILED DESCRIPTION

[0021] The exemplary embodiments will be described in detail hereinbelow with reference to the drawings. In the following description, the same numbers in different drawings represent the same or similar elements unless otherwise represented. The embodiments described in the following exemplary embodiments do not represent all the embodiments consistent with one or more embodiments of the present specification. Rather, they are merely examples of devices and methods consistent with some aspects of one or more embodiments of the present specification as detailed in the appended claims.

[0022] It should be noted that the steps of the corresponding method are not necessarily performed in the order shown and described in the present specification in other embodiments. In some other embodiments, the steps included in the method can be more or less than described in the present specification. In addition, a single step described in the present specification can be described as a plurality of steps in other embodiments, and a plurality of steps described in the present specification can be described as a single step in other embodiments.

[0023] The existing transcranial direct current stimulation (tDCS) technology in the home scene lacks professional guidance, has no effective electrode patch and head attachment effectiveness detection means, and cannot dynamically adapt to individual head shape and posture differences, resulting in low precision of user positioning of electrodes at core target points such as L-DLPFC, unstable efficacy, and safety risks such as skin stimulation.

[0024] Based on this, the present application solves the problems existing in the prior art based on artificial intelligence image recognition.

[0025] Embodiment one The present application aims to propose a transcranial direct current stimulation electrode positioning method based on artificial intelligence image recognition, specifically, referring to Figure 1 The method comprises the following technical steps: Step one, electrode patch preliminary detection stage To ensure the comfort, safety and durability of the device and the waterproofness of the device, the direct current path of the transcranial direct current stimulation (tDCS) device is metal electrode-conductive rubber electrode-disposable conductive gel electrode patch-user skin. To detect whether the disposable electrode patch is well adhered to the conductive rubber electrode, the cathode and the anode of the present application are both provided with sub-electrodes arranged in a semi-ring shape or a concentric circular shape. The system (a control system carried on the tDCS device, which can be simply referred to as a brain electrical system or a system, the control system carries the method of the present application, and the application program (APP) installed on the smart terminal also carries the method of the present application, and the two can work in cooperation) first applies a weak current between the two sub-electrodes inside a single electrode (cathode or anode), and detects the current amplitude and waveform stability. When the detected current reaches a stable preset threshold and the waveform is smooth, the system determines that the patch of the electrode has been correctly adhered, and prompts the user to proceed to the next step.

[0026] In the present embodiment, the surface of the electrode patch of the tDCS device is provided with one cathode and one anode (both are circular with a diameter of 30 mm), and the surface of the cathode and the anode is fixed with two semi-ring sub-electrodes (inner diameter 5 mm, outer diameter 8 mm, spacing 2 mm, and material is medical conductive silver paste); the device shell is made of ABS plastic, and two April Tag two-dimensional codes with a size of 15 mm x 15 mm and a depth of 0.1 mm are laser ablated on the outer side of the cathode and the anode 5 mm apart, for device pose recognition.

[0027] The current path adopts the structure of "metal electrode (medical stainless steel) → conductive rubber electrode (silicon-based conductive rubber, thickness 2 mm, resistivity ≤10 Ω·cm) → disposable conductive gel electrode patch (thickness 0.3 mm, water content 80%, conductivity ≥1 S / m) → patient skin", taking into account the comfort and conductivity.

[0028] The current detection adopts a current sensor (detection accuracy 0.01 mA) and a signal processing unit integrated in the tDCS device body; its functions are divided into two categories: one is to detect the current between the two sub-electrodes inside a single electrode (cathode or anode), and the other is to detect the current between the cathode and the anode; the detection data is transmitted in real time to the smart phone terminal or the smart terminal through Bluetooth, and the smart terminal here is the existing common smart phone, such as iphone15 series, Huawei mate 70 series, etc., which can call the camera after installing the application program to perform image recognition by using the algorithm built in the application program, that is, the complete stage of the subsequent step three, and the higher the configuration of the mobile phone, the better the effect. The present system adopts a downward compatible design, but the computing power of the terminal device will directly affect the time delay performance in the image recognition stage.

[0029] Preferably, as Figures 3-4As shown, the tDCS device has electrode shells 2 at both ends of the shell body 1, electrode patches are installed on the electrode shells 2, conductive rubber 3 is arranged on the electrode patches, the tDCS device is worn on the head of the user through a headband 4, the shell body 1 is further provided with a switch 5, an LED indicator 6, a charging port 7 and the like, which are all prior art, and the basic structure is not improved in the present application, but part of the structure is improved, and the method of the present application is carried.

[0030] Step two, whole head attachment detection phase When both electrode patches are detected, the system will prompt the user to wear the tDCS device on the head. In order to avoid the effective contact area being reduced due to the blocking of hair and the like, the system will perform whole attachment detection. At this time, the system outputs a specific detection waveform between the cathode and the anode, and monitors the current change trend between the two in real time. When it is monitored that the current presents a stable rising and finally maintains a stable direct current signal within a certain fluctuation range, the system determines that the device has been successfully attached to the skin of the user's head, and can enter the visual positioning phase.

[0031] Step three, AI visual intelligent positioning and calibration phase 1. Video data acquisition The user starts the matching application program on the intelligent terminal, and uses the front camera to shoot the facial image video stream in real time. Each frame of the video stream is transmitted to the image recognition algorithm module built in the application program in real time.

[0032] 2. OpenPose model key point labeling In this embodiment, the open source OpenPose artificial intelligence model is preferably used to process the video frames. Based on the convolutional neural network (CNN) and the pose estimation algorithm, the model can automatically and accurately identify and label the positions of 68 key points of the user's face and the bilateral pupils. These key points accurately define important anatomical landmarks of the user, such as the jawline (Jaw_0 to Jaw_16), eyebrows, eyes, nose, lips and the like. The use of the model ensures the accuracy of subsequent calculations and the robustness to different facial features of users.

[0033] Preferably, the facial key point labeling is the OpenPose model processing each frame of image, automatically identifying and labeling 68 facial key points (coordinate unit: pixel), including: Jawline: Jaw_0 (lower jaw left end point) to Jaw_16 (lower jaw right end point); Eyebrows: Eyebrow_Left_0 (left eyebrow left end point) to Eyebrow_Left_4 (left eyebrow inner point), Eyebrow_Right_0 (right eyebrow inner point) to Eyebrow_Right_4 (right eyebrow right end point); Eye: Eye_Left_0 to Eye_Left_6 (left eye contour), Eye_Right_0 to Eye_Right_6 (right eye contour), wherein Eye_Left_2 is the left eye pupil outside point, Eye_Right_2 is the right eye pupil outside point; Nose: Nose_0 to Nose_9 (nose contour); Lip: Lip_Outer_0 to Lip_Outer_11 (lip outside), Lip_Inner_0 to Lip_Inner_7 (lip inside), wherein Lip_Inner_6 is the upper lip midpoint; Simultaneously label the bilateral pupil positions (Pupil_Left_0 is the left pupil center, Pupil_Right_0 is the right pupil center).

[0034] 3. Algorithm calculates head electrode position Based on the real-time obtained key point two-dimensional coordinate data, the system calculates the target stimulation position of the tDCS electrode through a geometric calculation model, that is, the coordinates corresponding to the F3 point (left dorsolateral prefrontal cortex L-DLPFC) and F4 point (right dorsolateral prefrontal cortex R-DLPFC) in the international 10-20 electroencephalogram (EEG) system. The specific calculation steps are as follows: (1) Calculate the head size parameters (call the coordinate points output by the model, calculate the head size and set it as a parameter): jaw_0 = [keypoints_data.Jaw_0.X, keypoints_data.Jaw_0.Y]; jaw_16 = [keypoints_data.Jaw_16.X, keypoints_data.Jaw_16.Y]; eye_right_2 = [keypoints_data.Eye_Right_2.X, keypoints_data.Eye_Right_2.Y]; eye_left_2 = [keypoints_data.Eye_Left_2.X, keypoints_data.Eye_Left_2.Y]; jaw_distance = norm(jaw_0 - jaw_16); eye_distance = norm(eye_right_2 - eye_left_2); head_size_index = jaw_distance + eye_distance; This step is to adapt the algorithm to different head sizes of different users. First, a head size parameter needs to be calculated. For example, by calculating the distance between the two farthest points on the jaw line jaw_0 and jaw_16: jaw_distance, and the distance between the outer corners of the left and right eyes eye_left_2 and eye_right_2 eye_distance, and adding them together to get a comprehensive head size index head_size_index.

[0035] (2) Calculate the head tilt angle (determine the head tilt angle by the slope of the two pupils): pupil_left = [keypoints_data.Pupil_Left_0.X, keypoints_data.Pupil_Left_0.Y]; pupil_right = [keypoints_data.Pupil_Right_0.X, keypoints_data.Pupil_Right_0.Y]; head_tilt_slope = (pupil_right(2) - pupil_left(2)) / (pupil_right(1)- pupil_left(1)); This step is to correct the tilt of the user's head in the video frame. The system calculates the slope of the line connecting the left pupil pupil_left and the right pupil pupil_right head_tilt_slope as a parameter of the head tilt.

[0036] (3) Determine the position of the nasion (nasion) (determine the position of the nasion, i.e. AFz in the system, by the midpoint of the two inner eyebrow roots, which represents the midpoint of the front edge of the brain): eyebrow_left_4 = [keypoints_data.Eyebrow_Left_4.X, keypoints_data.Eyebrow_Left_4.Y]; eyebrow_right_0 = [keypoints_data.Eyebrow_Right_0.X, keypoints_data.Eyebrow_Right_0.Y]; nasion_position = (eyebrow_left_4 + eyebrow_right_0) / 2; This step takes the midpoint of the left medial eyebrow point, eyebrow_left_4, and the right medial eyebrow point, eyebrow_right_0, as a stable reference point on the facial midline, namely the nasion point, nasion_position. This position is approximately the AFz point in the EEG system.

[0037] (4) Infer the location of the midline of the forehead, Fz (determined by the extension of the line connecting the midpoint of the upper lip, lip_inner_6, and the nasion point, nasion_position (corrected via the head size parameter), which is located at the midpoint of F3 and F4): lip_inner_6 = [keypoints_data.Lip_Inner_6.X, keypoints_data.Lip_Inner_6.Y]; direction_vector = nasion_position - lip_inner_6; direction_vector = direction_vector / norm(direction_vector); Fz_distance = head_size_index * head_size_multiplier; Fz_position = nasion_position + direction_vector * Fz_distance; In this step, the system determines the direction vector of the facial midline, direction_vector, using the midpoint of the upper lip, lip_inner_6, and the nasion point, nasion_position. Then, it extends from the nasion point, nasion_position, along this direction vector upward by a distance, Fz_distance, which is positively related to the head size index, head_size_index, to determine the position of the key point Fz on the midline of the forehead, Fz_position. The Fz point is the midpoint of F3 and F4.

[0038] (5) Through the position of Fz, calculate L-DLPFC and R-DLPFC (i.e., the positions of F3 and F4 in the EEG system) via the head size parameter and the head slope parameter, which are the positions of the anode and cathode stimulation, respectively: distance_F3_F4 = head_size_index * head_size_multiplier * 0.8; direction_vector = [1, head_tilt_slope]; direction_vector = direction_vector / norm(direction_vector); F3_position = Fz_position + distance_F3_F4 * direction_vector; F4_position = Fz_position - distance_F3_F4 * direction_vector; This step is based on the determined Fz point position, the system constructs a direction vector related to the head tilt parameter head_tilt_slope, which is roughly parallel to the double pupil line. From the Fz point Fz_position, along the positive and negative directions of the vector respectively, a distance related to the head size index head_size_index is moved, which can calculate the coordinates of the left target point F3 and the right target point F4 respectively.

[0039] The above steps comprehensively utilize the geometric relationship of facial key points and the head shape parameter model to realize robust adaptation to individual differences (the test shows that the difference between the algorithm predicted point and the actual point is less than 5% of the head diameter), and has been verified on experimental participants of multiple countries, races and skin colors. The example output is shown in 2 (not representing the final UI effect, only an example of the embodiment of the application).

[0040] 4. Visualize real-time guidance The system superimposes the calculated F3 and F4 target position coordinates in the form of virtual markers (such as circles or crosses) on the video output interface of the intelligent terminal in real time. The user can intuitively see the deviation between the marker position and the actual tDCS device electrode, and manually fine-tune the position of the tDCS device according to the guidance on the screen until the device electrode and the virtual marker are completely coincident. As shown in the accompanying Figure 2 The algorithm shows good positioning effect on subjects of different races and facial features.

[0041] Step four, double positioning verification stage To further ensure the absolute accuracy of positioning, the embodiment also includes a double positioning verification step. On the shell of the tDCS device, two April Tag two-dimensional codes are pre-set (laser ablation method). After the user completes the positioning based on the visual guide, the camera of the smart terminal will recognize the two two-dimensional codes, and through an algorithm, the accurate physical position and rotation angle of the tDCS device in the three-dimensional space are calculated. The system compares the physical position information with the target position calculated based on the facial key points in real time, and verifies the accuracy of the actual center of the electrode patch and the target position calculated by the system, so as to realize double verification and ensure the best treatment effect. For example: The position deviation is less than or equal to 2mm, and the rotation angle deviation is less than or equal to 5°; if the threshold is met, the APP prompts "positioning success, start treatment"; if the deviation exceeds the threshold, the APP prompts "please adjust the device to align with the virtual marker", until the verification is qualified.

[0042] As a preferred embodiment, the present application can also directly replace OpenPose by using Google's MediaPipe Face Mesh model, which is a lightweight model optimized for mobile terminals, based on CNN architecture, and can output 468 3D facial key points (far exceeding 68 basic points, including pupil, eyebrow, nose wing, etc. Subdivision structure) in real time, with a positioning accuracy of ±0.8mm; It is robust to head tilt (±45°) and partial occlusion (such as glasses, bangs), and is suitable for non-standard poses of users in a home environment; it supports direct output of three-dimensional coordinates (x, y, z) of key points. Or the existing open source Dlib Face Landmark Predictor model, which is a classic facial key point detection model based on HOG features + support vector machine (SVM), can stably output 68 two-dimensional facial key points with a positioning accuracy of ±0.6mm, and the training data covers multiple skin colors and age groups. By supplementing the pupil detection position (such as using the MediaPipe Face Mesh model), OpenPose can be seamlessly replaced. The actual situation can be selected.

[0043] In the prior art, AI vision models (such as OpenPose, MediaPipe) are mainly used in general scenarios such as face recognition and motion capture, current detection technology is mainly used in circuit fault diagnosis, and tDCS electrode positioning relies on manual operation in a medical environment. These technologies have developed independently in their respective fields, and there is no suggestion in the prior art to integrate the three technologies for "home tDCS electrode positioning".

[0044] The core innovation of the present application is: 1. Precisely anchoring AI vision technology to medical anatomical needs: Rather than simply using facial key point detection, this approach specifically identifies anatomical landmarks (such as the inner brow point, pupil, and jawline) that are strongly correlated with EEG leads (F3 / F4 / Fz). A geometric inference model based on these key points is then designed to achieve precise mapping from facial features to brain region targets (this mapping relationship is not publicly documented in existing technologies). 2. Progressive synergy between current feedback and visual positioning: Sub-electrode current detection is first used to eliminate the fundamental error of "the patch is not firmly attached." The "overall head attachment effectiveness" is then determined by the current trends of the cathode and anode. Finally, AI visual positioning is used. This "hardware detection as a foundation for visual positioning" logic solves the problem of "invalid positioning due to device not being attached" when using AI vision alone, and is a creative combination of the two technologies. 3. Scenario-based design of a dual verification mechanism: By comparing the April Tag QR code with key facial points, the "device posture" is linked to the "head anatomical reference." This overcomes the limitation of AI vision that can only locate the target but cannot verify the actual position of the device, forming a "positioning-verification" closed loop. This mechanism has not been used in existing tDCS or AI vision fields.

[0045] In summary, the tDCS electrode intelligent positioning method based on AI visual recognition and current feedback proposed in the present invention uses OpenPose to accurately mark head feature points in real time, combines rigorous algorithm reasoning and a visual user guidance interface, and constructs a complete technical closed loop from "patch detection → head attachment → target positioning → real-time verification", which solves the "precision-safety-ease of use" balance problem that has long been unresolved by existing technologies. It can realize high-precision and adaptive automatic positioning and attachment of electrodes in home scenarios, significantly improving the accuracy and convenience of transcranial direct current stimulation equipment.

[0046] Example 2 Based on the same concept, the present invention also proposes a transcranial direct current stimulation electrode positioning device based on artificial intelligence image recognition, comprising: The electrode assembly includes a cathode and an anode, and sub-electrodes and a conductive medium are provided on the cathode and the anode; A current detection module is used to detect the current between the sub-electrodes and the current between the cathode and anode to evaluate the patch adhesion and the overall head attachment status; An image acquisition module, used to collect real-time image data of the user's face; An artificial intelligence processing module is configured to identify preset key points of a user's face in real-time image data, and calculate target position coordinates of at least one transcranial direct current stimulation electrode on the user's head based on a geometric calculation model and the preset key points, wherein the preset key points include feature points of a user's jawline, eyebrows, eyes, nose and lip region. A visual guidance module is configured to superimpose the target position coordinates on a video output interface in real time to guide the user to place the transcranial direct current stimulation electrode at the target position. A positioning verification module is configured to identify positioning markers on a device surface, acquire pose information of a transcranial direct current stimulation electrode device relative to the head, and perform real-time verification with the face key point information.

[0047] Embodiment Three The embodiment also provides an electronic device, which refers to Figure 5 The electronic device includes a memory 404 and a processor 402, the memory 404 stores a computer program, and the processor 402 is configured to run the computer program to perform the steps in any of the above method embodiments.

[0048] Specifically, the processor 402 can include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0049] The memory 404 can include a mass storage that stores data or instructions. For example, and without limitation, the memory 404 can include a Hard Disk Drive (HDD), a floppy disk drive, a Solid State Drive (SSD), a flash drive, a Compact Disc Read Only Memory (CD-ROM), a magneto-optical disk, a magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. The memory 404 can be removable and / or non-removable (or fixed) as appropriate. The memory 404 can be internal or external as appropriate. In particular embodiments, the memory 404 is a Non-Volatile memory. In particular embodiments, the memory 404 includes a Read-Only Memory (ROM) and a Random Access Memory (RAM). The ROM can be a mask-programmed ROM, a Programmable ROM (PROM), an Erasable PROM (EPROM), an Electrically EPROM (EEPROM), an Electrically Alterable ROM (EAROM), or a FLASH memory, or a combination of two or more of these, as appropriate. The RAM can be a Static Random-Access Memory (SRAM) or a Dynamic Random Access Memory (DRAM), which can be a Fast Page Mode Dynamic Random Access Memory (FPMDRAM), an Extended Data Output Dynamic Random Access Memory (EDODRAM), a Synchronous Dynamic Random-Access Memory (SDRAM), or the like, as appropriate.

[0050] The memory 404 can be used to store or buffer various data files needed for processing and / or communication, and possible computer program instructions executed by the processor 402.

[0051] The processor 402 implements any one of the above-mentioned artificial intelligence image recognition-based transcranial direct current stimulation electrode positioning methods by reading and executing the computer program instructions stored in the memory 404.

[0052] Optionally, the above-mentioned electronic device can further include a transmission device 406 connected with the processor 402 and an input / output device 408 connected with the processor 402.

[0053] The transmission device 406 can be used to receive or send data via a network. Specific examples of the network can include a wired or wireless network provided by a communication provider of the electronic device. In one example, the transmission device includes a network adapter (NIC) which can be connected with other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 406 can be a radio frequency (RF) module which is used to communicate with the Internet in a wireless manner.

[0054] The input / output device 408 is used to input or output information.

[0055] Embodiment Four The embodiment also provides a readable storage medium, and the readable storage medium stores a computer program. The computer program includes program codes for controlling a process to execute the process. The process includes the artificial intelligence image recognition-based transcranial direct current stimulation electrode positioning method according to Embodiment One.

[0056] It should be noted that specific examples in the embodiment can refer to examples described in the above-mentioned embodiments and optional implementation manners, and the embodiment will not be described here again.

[0057] Generally, various embodiments can be implemented in hardware or special-purpose circuitry, software, logic or any combination thereof. Some aspects of the application can be implemented in hardware, while other aspects can be implemented by firmware or software executed by a controller, microprocessor or other computing device, but the application is not limited thereto. Although various aspects of the application can be illustrated and described as block diagrams, flow charts, or using some other pictorial representation, it is well understood that these blocks, apparatus, systems, techniques or methods described herein can be implemented in hardware, software, firmware, special purpose circuits or logic, general purpose hardware or controller or other computing devices, or some combination thereof.

[0058] Embodiments of the application can be implemented by computer software executable by a data processor of the mobile device such as in the processor entity, or by hardware, or by a combination of software and hardware. Computer software or program, also called program product, including software routines, applets and / or macros, can be stored in any apparatus-readable data storage medium and they include program instructions to implement certain tasks. The program product can include one or more computer-executable components tangibly embodied in a computer- readable medium, when executed, for implementing one or more embodiments of the present application. The one or more computer-executable components can be one or more of a procedure, a function, a subprogram, a plugin, a module, a software

[0059] Those skilled in the art should clearly understand that each technical feature in the above embodiments can be combined with any other technical feature, and for the sake of brevity, not all possible combinations of the technical features in the above embodiments are described, however, as long as the combinations of the technical features do not exist in contradiction, they should be considered within the scope of the present disclosure.

[0060] The above embodiments only express several implementation manners of the present application, and the description is relatively specific and detailed, but it should not be understood as a limitation on the scope of the present application. It should be pointed out that for those skilled in the art, some modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A transcranial direct current stimulation electrode positioning method based on artificial intelligence image recognition, using a transcranial direct current stimulation device, characterized in that: The following steps are involved: Acquire real-time image data of the user's face; Using an artificial intelligence facial recognition model to process the real-time image data, identify and output the coordinates of a plurality of preset key points on the user's face, the preset key points including feature points of the user's jawline, eyebrows, eyes, nose, and lips; Based on the coordinates of the multiple preset key points, the target position coordinates of at least one transcranial direct current stimulation electrode on the user's head are calculated using a preset geometric calculation model; On the display interface, visual guidance information is generated according to the target position coordinates to guide the user to place the transcranial direct current stimulation electrode at the target position.

2. The method for positioning transcranial direct current stimulation electrodes based on artificial intelligence image recognition according to claim 1, wherein: The step of calculating the target position coordinates using the geometric dead reckoning model includes: Calculating a head size parameter for characterizing a head size of the user according to a distance between at least two key points; Determining a head tilt parameter for characterizing a head tilt state based on relative positions of key pupil points of the user's two eyes; Determine the first reference point on the facial midline based on the key points of the eyebrow or nose; Calculating a second reference point on the midline of the skull based on the first reference point, other key points on the face, and the head size parameters; The target position coordinates are calculated by combining the second reference point, the head size parameter, and the head tilt parameter.

3. The method for positioning transcranial direct current stimulation electrodes based on artificial intelligence image recognition according to claim 1, wherein: The target location is an area corresponding to the left dorsolateral prefrontal cortex or the right dorsolateral prefrontal cortex.

4. The method for positioning transcranial direct current stimulation electrodes based on artificial intelligence image recognition according to claim 1, wherein: Also includes a double positioning verification step: Obtaining real-time spatial position information of the transcranial direct current stimulation device through a machine-readable code set on the housing of the image recognition device; The spatial posture information is compared with the target position calculated through facial key points to confirm that the actual placement position of the transcranial direct current stimulation electrode is consistent with the target position.

5. The method for positioning transcranial direct current stimulation electrodes based on artificial intelligence image recognition according to claim 4, wherein: The machine-readable code is an April Tag QR code.

6. The method for positioning transcranial direct current stimulation electrodes based on artificial intelligence image recognition according to claim 1, wherein: The artificial intelligence facial recognition model is an OpenPose model based on a convolutional neural network.

7. A method for transcranial direct current stimulation electrode positioning based on artificial intelligence image recognition according to any one of claims 1 to 6, characterized in that: Before acquiring real-time image data of the user's face, at least one of the following electrical detection steps is also included: Preliminary electrode patch testing: By detecting the current parameters between the sub-electrodes inside the electrode, it is determined whether the disposable electrode patch is well attached to the transcranial direct current stimulation electrode; Overall head attachment detection: By applying a current of a preset waveform between the cathode electrode and the anode electrode and performing a detection, it is determined whether the transcranial direct current stimulation device is well attached to the user's head.

8. A transcranial direct current stimulation electrode positioning device based on artificial intelligence image recognition, characterized in that: include: An electrode assembly, comprising a cathode and an anode, wherein sub-electrodes and a conductive medium are provided on the cathode and the anode; A current detection module is used to detect the current between the sub-electrodes and the current between the cathode and anode to evaluate the patch adhesion and the overall head attachment status; An image acquisition module, used to collect real-time image data of the user's face; An artificial intelligence processing module for identifying preset key points on the user's face in real-time image data and calculating the target position coordinates of at least one transcranial direct current stimulation electrode on the user's head using a geometric calculation model based on the coordinates of the preset key points, including feature points in the user's jawline, eyebrows, eyes, nose, and lips. A visual guidance module is used to superimpose the target position coordinates on the video output interface in real time to guide the user to place the transcranial direct current stimulation electrode at the target position; The positioning verification module is used to identify the positioning marks on the surface of the device, obtain the position information of the transcranial direct current stimulation device relative to the head, and perform real-time verification with the facial key point information.

9. An electronic device comprising a memory and a processor, characterized in that: The memory stores a computer program, and the processor is configured to run the computer program to execute the transcranial direct current stimulation electrode positioning method based on artificial intelligence image recognition according to any one of claims 1 to 7.

10. A readable storage medium, characterized in that: The readable storage medium stores a computer program, which includes a program code for controlling a process to execute a process, and the process includes the transcranial direct current stimulation electrode positioning method based on artificial intelligence image recognition according to any one of claims 1 to 7.

Images