Medical data processing method and system for reversible adhesion of hydrogel brain electrodes

Abstract

The application discloses a kind of hydrogel brain electrode reversible adhesion medical data processing method and system, it is related to medical data processing technical field, the specific steps of this method are as follows: first, reversible adhesion electrode is attached to patient brain, and dynamic data set is formed by cyclically collecting multimodal data;Then the data is preprocessed and fused to generate a correlation graph;Then, based on the transfer learning model, the diagnosis result and the prescription are output;Then, the state is detected, the parameter is determined and the drug is delivered by performing collaborative intervention;Finally, the data after intervention is collected according to the node, and the algorithm is optimized and updated to form a cyclic optimization link.The application integrates reversible adhesion electrode and multimodal acquisition technology, constructs brain function-molecular mechanism correlation graph, breaks information limitation, and improves disease recognition accuracy;Through transfer learning and other methods, personalized brain electrical feature map is constructed, collaborative intervention and closed-loop feedback optimization are realized relying on various algorithms, traditional intervention problems are solved, and precise and dynamic development of neural regulation is promoted.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of medical data processing, in particular to a medical data processing method and system for reversible adhesion of hydrogel brain electrodes. BACKGROUND

[0002] With the rapid development of brain science research and neural regulation technology, the diagnosis and treatment of brain diseases gradually move towards precision and individualization. As a core tool for obtaining brain electrophysiological signals, semi-dry brain electrodes play a key role in the clinical research and treatment of brain diseases such as epilepsy and cognitive impairment, and in the field of brain-computer interface technology. At the same time, a single brain electrical signal has been difficult to meet the diagnosis needs of complex brain diseases, and in clinical practice, it is gradually necessary to combine physiological state, molecular expression and behavioral characteristics and other multi-dimensional information to fully reveal the pathological mechanism of the disease. In addition, the adhesion performance of the brain electrode directly affects the stability and continuity of signal acquisition. The electrical adhesion property can realize the rapid and stable fitting of the electrode and the scalp, reduce the pressure damage of mechanical fixation to the patient's brain skin, support repeated acquisition and intervention operation. Under this background, how to realize the efficient acquisition and accurate fusion of multi-modal data, combined with the electrical adhesion type semi-dry brain electrode to construct a complete link from data processing to individualized intervention, has become an important direction to promote the clinical transformation of neural regulation technology.

[0003] Traditional brain electrode technology has obvious limitations: although wet brain electrodes can reduce contact impedance and obtain high-quality brain electrical signals, they need to be coated with conductive paste, which is cumbersome to use and inconvenient to clean, and can easily cause skin discomfort or allergies; dry brain electrodes are easy to operate, but the contact with the scalp is unstable, the signal noise is large, and the signal acquisition quality is poor. The semi-dry brain electrode combines the advantages of both, but the existing semi-dry brain electrode mostly adopts mechanical fixation, which can easily cause pressure on the scalp and affect comfort, and the fixation effect is greatly affected by factors such as movement, and lacks an efficient adhesion scheme based on electrical response characteristics. In terms of data processing, traditional methods mostly rely on single modal data for analysis, which is difficult to effectively integrate brain electrical, physiological, molecular and behavioral data, resulting in incomplete disease characteristics extracted, inability to establish precise correlation between brain function and molecular mechanism, and thus affecting the accuracy of the diagnosis result. In the model training link, the traditional model is often trained with a fixed data set and directly applied, lacking individualized adjustment and dynamic incremental update capability for individual differences, and being difficult to adapt to the dynamic changes of the patient's condition. In the intervention execution, the traditional scheme mostly separately uses transcranial electrical stimulation or drug therapy, lacks the collaborative optimization design of the two, and lacks real-time effect feedback and parameter adjustment mechanism, resulting in insufficient pertinence of the intervention scheme, and difficulty in balancing efficacy and safety.

[0004] The existing adhesion technology for semi-dry brain electrodes presents significant challenges: balancing stability and comfort is difficult with mechanical fixation methods, while conventional adhesion materials lack electroresponsiveness, making it impossible to quickly regulate the adhesion state via external signals. Furthermore, the adhesion effect tends to decline after repeated use, affecting not only the continuity of data acquisition but also potentially causing signal distortion due to improper fixation. Therefore, developing semi-dry brain electrodes based on electroresponsiveness and employing a dual-network interpenetrating conductive hydrogel with a composite PVA and PAM network structure, achieving rapid and stable electroadhesion through the application of a continuous DC voltage, while simultaneously integrating multimodal data processing and closed-loop intervention technologies, is key to overcoming the current technological bottlenecks. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a medical data processing method and system for reversibly adhered hydrogel brain electrodes. This method collects multimodal data, including EEG and physiological parameters, using semi-dry brain electrodes. After preprocessing and cross-modal fusion, it constructs an EEG feature atlas using transfer learning and personalized incremental training, and outputs disease diagnosis and personalized neuromodulation prescriptions. The system includes modules for data acquisition, preprocessing and fusion, model training, collaborative intervention, and closed-loop optimization to achieve precision medicine and dynamic adjustment.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: On one hand, a method for medical data processing using reversible adhesion of hydrogel brain electrodes, the specific steps of which are as follows:

[0007] S100, Multimodal Dynamic Acquisition: A semi-dry brain electrode integrating a micro-programmable transcranial AC stimulation array is attached to the target functional area of ​​the patient's brain. The core conductive layer of the semi-dry brain electrode is a conductive hydrogel with a dual-network interpenetrating structure composed of PVA network and PAM network. By applying a continuous DC voltage to the conductive hydrogel, the electro-response characteristics of the material cause the conductive hydrogel to undergo physical changes and generate adhesion force, achieving rapid and stable electro-adhesion between the brain electrode and the scalp. Through the electrode and supporting acquisition equipment, EEG signals, physiological parameters, molecular expression imaging data and behavioral feature data are cyclically acquired to form a multi-dimensional dynamic dataset.

[0008] S200, Data Preprocessing and Fusion: Denoising, normalization, and spatial registration preprocessing are performed on the collected multimodal data respectively. A cross-modal spatiotemporal attention fusion algorithm is used to extract key features of each modality and weighted concatenation to generate cross-modal joint feature vectors and brain function-molecular mechanism association maps.

[0009] The specific steps of the brain function-molecular mechanism correlation map are: defining the brain function partitions of prefrontal lobe, parietal lobe, temporal lobe, occipital lobe and hippocampus based on brain anatomy, and determining the spatial coordinate range of each partition in combination with brain MRI image; extracting the expression concentration, spatial distribution range and expression intensity gradient data of specific molecular markers in each brain function partition from the preprocessed molecular expression imaging data; calculating the correlation coefficient of the brain electrical core features of each brain function partition and the expression data of the corresponding partition molecular markers, and quantifying the correlation strength of the two; taking the brain function partition as the horizontal axis and the molecular marker type as the vertical axis, constructing a two-dimensional map framework, and filling the correlation strength into the corresponding coordinate position in the form of numerical value or color gradient; incorporating the spatial correlation matrix information in the cross-modal spatio-temporal attention fusion algorithm to correct the correlation weight of each partition and molecular marker, and forming a complete and quantitatively accurate brain function-molecular mechanism correlation map;

[0010] S300, model training and prescription generation: initializing the model framework based on transfer learning, using personalized adaptive incremental training algorithm, iteratively updating model parameters using dynamic data set, constructing personalized electroencephalogram feature map, outputting disease diagnosis result and personalized neuroregulation prescription;

[0011] The specific way of outputting the disease diagnosis result and the personalized neuroregulation prescription is: inputting the cross-modal joint feature vector, the personalized electroencephalogram feature map and the brain function-molecular mechanism correlation map into the model optimized by the personalized adaptive incremental training algorithm; the model strengthens the correlation mapping of individual-specific electroencephalogram features and molecular expression rules through the feature extraction network, and outputs the disease diagnosis result after classification calculation by the full connection layer; based on the pathological mechanism correlation in the diagnosis result, the electroencephalogram abnormal feature and the molecular marker expression rule of the corresponding brain function partition are extracted; combined with the brain function partition positioning result, the core parameters of the personalized neuroregulation prescription are determined: the stimulation target is the brain function partition corresponding to the electroencephalogram abnormal feature, the stimulation frequency matches the electroencephalogram abnormal wave band characteristics, the stimulation intensity is adapted according to the individual skin impedance and electroencephalogram signal amplitude, and the stimulation duration is related to the molecular marker expression concentration; integrating the above parameters to form a complete personalized neuroregulation prescription, which is output to the system terminal synchronously with the disease diagnosis result, providing instruction support for subsequent collaborative intervention;

[0012] S400, collaborative intervention execution: detecting the adhesion state through the sensor integrated with the electrode, determining the optimal neuroregulation parameter and drug dose by using the collaborative intervention parameter optimization algorithm, and sequentially executing transcranial electrical stimulation and drug delivery, continuously collecting monitoring data throughout the process;

[0013] S500, closed-loop optimization update: collect data after intervention again at the preset time node, supplement to the dynamic data set, calculate the parameter update amount using the closed-loop effect feedback optimization algorithm, adjust the model parameters, intervention parameters and collection frequency, and form a cycle optimization link.

[0014] Further, in the S100, in the multi-modal dynamic acquisition, the conductive hydrogel of the semi-dry electrode is formed by compounding PVA network and PAM network at a mass ratio of 2:1-3:1 to form a double-network interpenetrating structure, adding an electric response characteristic enhancer accounting for 3%-6% of the total mass of the hydrogel, and shaping after 3-5 freeze-thaw cycles, the tensile rate is ≥1200%, the tensile strength is ≥280kPa, the electric adhesion reversible times are ≥15 times, and the adhesion is released within 30s after power-off; the target functional area of the patient's brain includes the parietal lobe and hippocampus of Alzheimer's disease patients, the temporal lobe and frontal lobe of epilepsy patients, and the lesion associated area determined by brain MRI positioning; the matching acquisition equipment includes Ag / AgCl electroencephalogram sensor, miniature piezoresistive pulse sensor, NTC thermistor temperature sensor, CdSe / ZnS core-shell structure quantum dot marker probe, near-infrared fluorescence imaging system and wearable acceleration sensor.

[0015] Further, in the S100, in the multi-modal dynamic acquisition, the collected electroencephalogram signal includes brain electrical activity signal, the physiological parameter includes pulse frequency and skin temperature, the molecular expression imaging data includes expression concentration data and spatial distribution data of specific molecular markers in the brain, and the behavior characteristic data includes walking speed, limb tremor frequency and limb activity amplitude data.

[0016] Further, in the S200, in the data preprocessing and fusion, the mathematical expression of the cross-modal spatio-temporal attention fusion algorithm is: wherein, is the final output cross-modal joint feature vector, corresponding to brain electrical activity, molecular imaging, physiology and behavior, is the modal reference weight, is the modal reliability factor, is the modal feature vector, is the spatial correlation matrix, is the time series attenuation coefficient, is the collection time interval.

[0017] Further, in the model training and prescription generation, the specific steps of constructing the individualized electroencephalogram feature map of S300 are: screening electroencephalogram related core features from the cross-modal joint feature vector, including electroencephalogram time domain features, frequency domain features and the association features of electroencephalogram signals and molecular marker expression; based on the model output initialized by transfer learning, combined with brain anatomical structure partition, a brain function partition map framework is established, which contains prefrontal lobe, parietal lobe, temporal lobe, occipital lobe and hippocampus; the feature values corresponding to the electroencephalogram data obtained in each adhesion-acquisition cycle are filled into the map framework according to the brain function partition to form an initial individualized electroencephalogram feature map; the feature weight and correlation strength of each brain function partition are corrected by using the model parameters updated by the individualized adaptive incremental training algorithm, and the representation of individual specific electroencephalogram features is strengthened; the map is updated synchronously according to the number of adhesion-acquisition cycles, the dynamic change trajectory of the electroencephalogram features of each partition at different time nodes is recorded, and a complete individualized electroencephalogram feature map is formed.

[0018] Further, in the model training and prescription generation, the specific content of the model framework initialized by transfer learning of S300 is: the pre-training data set is selected as the PhysioNet electroencephalogram public data set; the basic model selects ResNet-18 network, which contains 16 convolutional layers, 4 pooling layers and 3 fully connected layers, wherein the convolutional layers are used for feature extraction, and the fully connected layers are used for classification output; the parameters of the 16 convolutional layers and the 4 pooling layers in the pre-trained model trained by the public data set are retained as the initial weights of feature extraction; the original fully connected layer is replaced by 3 layers of adaptive structure, the first layer contains 1024 neurons, the second layer contains 512 neurons, and the third layer contains 64 neurons, and the output layer adopts Softmax activation function, which adapts to the classification task of 3 disease types and 3 risk levels; the He initialization method is used to initialize the parameters of the replaced fully connected layer, and the initial learning rate is set to 0.001, so as to complete the initialization of the model framework by transfer learning.

[0019] Further, in the model training and prescription generation, the model parameter update formula of the individualized adaptive incremental training algorithm of S300 is: wherein, is the model parameter after the first acquisition, is the model parameter after the first acquisition, is the individualized forgetting factor, which is dynamically adjusted according to the stability of the patient's historical data, is the novelty coefficient of new data, which is calculated based on the difference between new data and old data, is the adaptive learning rate, is the cross-entropy loss function gradient, is the final output cross-modal joint feature vector, For brain function-molecular mechanism correlation atlas, For diagnostic label, For the Effective data volume of the first acquisition, For data volume baseline value.

[0020] Further, the S400, in the cooperative intervention execution, the mathematical expression of the cooperative intervention parameter optimization algorithm is: The constraint condition is kPa, Wherein, The final optimal parameter combination to be found, Is a mathematical operator, looking for the value of the independent variable that makes the function reach the maximum value ), The optimal neuromodulation parameter, The optimal stimulation frequency, The optimal stimulation intensity, The optimal stimulation duration, The optimal drug dose, , The weight coefficient, The abnormal feature improvement of electroencephalogram, The total duration of intervention, The physiological parameter fluctuation, The molecular marker concentration reduction, The actual adhesion pressure of hydrogel electrode, The skin temperature of the electrode attachment area.

[0021] Further, in the S500, in the closed-loop optimization update, the mathematical expression of the closed-loop effect feedback optimization algorithm is: Wherein, The parameter update amount, The optimization coefficient, The cooperative effect index is determined based on the cooperative change degree of multi-modal data before and after intervention, The cross-modal joint feature vector after intervention, The cross-modal joint feature vector before intervention.

[0022] On the other hand, a power distribution system based on the above power big data automatic reasoning platform, the system comprises:

[0023] Data acquisition module: composed of semi-dry electrodes and matching acquisition equipment, used for attaching to the target functional area of the patient's brain, cyclically acquiring electroencephalogram signals, physiological parameters, molecular expression imaging data and behavior characteristic data, forming a multi-dimensional dynamic data set;

[0024] ​Data preprocessing and fusion module: for performing denoising, normalization, spatial registration preprocessing on the collected multi-modal data, and generating a cross-modal joint feature vector through a cross-modal spatio-temporal attention fusion algorithm;

[0025] Model training and prescription generation module: for initializing the model framework based on transfer learning, iteratively updating the model parameters through a personalized adaptive incremental training algorithm, constructing a personalized electroencephalogram feature map, and outputting a disease diagnosis result and a personalized neuroregulation prescription;

[0026] Collaborative intervention execution module: for detecting the electrode adhesion state through a sensor, determining the optimal neuroregulation parameters and drug dosage based on a collaborative intervention parameter optimization algorithm, executing transcranial electrical stimulation and drug delivery, and continuously collecting monitoring data;

[0027] Closed-loop optimization and update module: for collecting post-intervention data at preset time nodes, supplementing to the dynamic data set, adjusting the model parameters, intervention parameters and collection frequency through a closed-loop effect feedback optimization algorithm, and forming a cycle optimization link.

[0028] Compared with the prior art, the medical data processing method and system with reversible adhesion of hydrogel brain electrodes have the following beneficial effects:

[0029] Firstly, the present application integrates reversible adhesion electrodes and multi-modal dynamic acquisition technology to comprehensively obtain electroencephalogram signals, physiological parameters, molecular expression imaging data and behavior characteristic data, which are preprocessed by denoising, normalization and spatial registration, and then key features of each modality are extracted and weighted spliced using a cross-modal spatio-temporal attention fusion algorithm to construct a brain function-molecular mechanism correlation map.

[0030] Secondly, the application initializes a model framework through transfer learning, iteratively updates model parameters by combining a personalized adaptive incremental training algorithm, constructs a dynamically evolving personalized electroencephalogram feature map, accurately outputs disease diagnosis results and customized neuromodulation prescriptions, relies on a collaborative intervention parameter optimization algorithm to realize the collaborative cooperation of transcranial electrical stimulation and drug delivery, simultaneously monitors the electrode adhesion state and intervention process data in real time through sensors, continuously integrates the post-intervention data to adjust the model parameters, intervention scheme and acquisition frequency by means of a closed-loop effect feedback optimization algorithm, forms a whole-process cyclic optimization link, this combination mode of personalization and dynamic optimization not only ensures that the intervention scheme is in line with the individual pathological characteristics and body state, but also can respond to the intervention effect changes in real time, improves the pertinence and safety of the intervention, effectively solves the problems of lack of personalization and adjustment lag in the traditional intervention scheme, promotes the development of neuromodulation treatment towards precision and dynamic direction, and significantly improves the treatment effect and patient experience.

[0031] Other advantages, objects, and features of the application will be set forth in part in the following specification taken in conjunction with the accompanying drawings, and in part will become apparent to those skilled in the art from a consideration of the following specification and drawings. It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application as claimed. BRIEF DESCRIPTION OF DRAWINGS

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creating any creative labor.

[0033] Figure 1 The framework diagram of the hydrogel brain electrode reversible adhesion medical data processing method;

[0034] Figure 2 The flowchart of the hydrogel brain electrode reversible adhesion medical data processing system. DETAILED DESCRIPTION

[0035] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined application purpose, the specific embodiments, structures, features and effects according to the present application will be described in detail below in combination with the drawings and preferred embodiments.

[0036] Embodiment one:

[0037] Medical data processing embodiment for Alzheimer's disease patients.

[0038] Multi-modal dynamic acquisition: A semi-dry EEG electrode was prepared by mixing PVA network and PAM network at a mass ratio of 3:1, and adding an electric response characteristic enhancer accounting for 5% of the total mass of the hydrogel. After 4 freeze-thaw cycles, the electrode was formed. The electrode has a tensile strength of 1580% and a tensile strength of 320 kPa. The electric adhesion is reversible for more than 15 times, and the adhesion force is completely released within 25 seconds after power-off. The electrode integrates a microprogrammable transcranial alternating current stimulation array. The high tensile strength and tensile strength of the electrode can adapt to the dynamic changes of the brain profile, and the multiple reversible adhesion characteristics support long-term cyclic acquisition, avoiding frequent replacement of the electrode to stimulate or discomfort the scalp of the patient. According to the lesion-related area of the patient's brain MRI positioning, the electrode is precisely attached to the parietal lobe and the hippocampal region, which are two target functional areas closely related to the pathological changes of Alzheimer's disease. Through the matching Ag / AgCl EEG sensor, micro piezoresistive pulse sensor, NTC thermistor temperature sensor, CdSe / ZnS core-shell structure quantum dot marker probe, near-infrared fluorescence imaging system and wearable acceleration sensor, the data are cyclically acquired: the Ag / AgCl EEG sensor captures the brain electrical activity signal, reflecting the activity state of the brain neurons; the micro piezoresistive pulse sensor and the NTC thermistor temperature sensor monitor the pulse frequency, skin temperature and other physiological parameters to master the patient's basic physical condition; the CdSe / ZnS core-shell structure quantum dot marker probe combined with the near-infrared fluorescence imaging system obtains the expression concentration and spatial distribution data of the brain specific molecular marker, revealing the pathological changes at the molecular level; the wearable acceleration sensor records the walking speed, limb tremor frequency and limb activity amplitude and other behavior characteristic data, reflecting the actual functional performance of the patient. The multi-dimensional data together form a dynamic data set, providing comprehensive and three-dimensional information support for subsequent diagnosis and intervention. The electrode attachment strictly follows the lesion-related area of the brain MRI positioning, ensuring that the parietal lobe and hippocampal region are accurately covered in the pathological related area; the integrated microprogrammable transcranial alternating current stimulation array is ready for synchronization, providing hardware support for the transcranial electrical stimulation segment in the subsequent collaborative intervention, and recording the acquisition time interval of each modality data during the acquisition process, providing timing parameters for cross-modality spatio-temporal attention fusion algorithm, such as Figure 1 The hydrogel EEG electrode reversible adhesion medical data processing method framework diagram shown in the figure directly presents the whole process of multi-modal dynamic acquisition-data preprocessing and fusion-model training and prescription generation-collaborative intervention execution-closed loop optimization update: first, multi-dimensional data is collected by the reversible adhesion electrode, and after processing and fusion, the model is trained to output diagnosis and prescription, then the collaborative intervention is executed, and finally the parameters are adjusted according to the data after the intervention, forming a complete link of cyclic optimization.

[0039] Data preprocessing and fusion: For the four types of modal data collected, including electroencephalogram (EEG), physiological, molecular expression imaging and behavioral characteristics, preprocessing such as denoising, normalization and spatial registration is performed. Denoising removes irrelevant signals such as environmental electromagnetic interference and electromyographic interference to ensure the purity of EEG data. Normalization unifies the dimensions of different modal data to avoid affecting the subsequent feature fusion effect due to data range differences. Spatial registration accurately corresponds each modal data to the brain anatomy structure to ensure the consistency of data in spatial position and lay an accurate foundation for cross-modal data correlation analysis. Based on the brain anatomy structure, the brain function partitions of prefrontal lobe, parietal lobe, temporal lobe, occipital lobe and hippocampus are defined, and the spatial coordinate ranges of each partition are determined combined with the brain MRI image. From the preprocessed molecular expression imaging data, the expression concentration, spatial distribution range and expression intensity gradient data of specific molecular markers in each brain function partition are extracted. By calculating the correlation coefficient between the EEG core features of each brain function partition and the expression data of the corresponding molecular markers, the correlation strength of the two is quantified, and the corresponding relationship between brain function changes and molecular mechanisms is clearly presented, providing a reference for understanding the molecular root of brain function abnormalities. Taking the brain function partition as the horizontal axis and the molecular marker type as the vertical axis, a two-dimensional graph framework is constructed, and the correlation strength is filled in the form of numerical value or color gradient. Then, the spatial correlation matrix information is integrated to correct the correlation weight, forming a complete and quantitative brain function-molecular mechanism correlation graph. Subsequently, the cross-modal spatio-temporal attention fusion algorithm is used, which can automatically identify and highlight the key features related to disease diagnosis in each modality, and weaken irrelevant information. By weighting and splicing the key features of each modality, a cross-modal joint feature vector is generated. The mathematical expression of the cross-modal spatio-temporal attention fusion algorithm is: wherein, is the final output cross-modal joint feature vector, corresponds to the four modalities of EEG, molecular imaging, physiology and behavior, is the modal reference weight, is the modal reliability factor, is the feature vector of each modality, is the spatial correlation matrix, is the time series decay coefficient, is the collection time interval, integrating multi-dimensional information to avoid the limitations of single modal data and providing comprehensive feature support for subsequent model training.

[0040] Model Training and Prescription Generation: The PhysioNet EEG public dataset was selected as the pre-training dataset. The basic model used was a ResNet-18 network containing 16 convolutional layers, 4 pooling layers, and 3 fully connected layers. Transfer learning was performed to initialize the model framework: The parameters of the 16 convolutional layers and 4 pooling layers trained on the public dataset in the pre-training model were retained. These parameters contain general features of brain electrical signal processing, which can reduce the computational cost of training the model from scratch, speed up the initialization, and improve the model's basic ability to extract EEG features. The original fully connected layers were replaced with a 3-layer adaptive structure. The first layer contains 1024 neurons, the second layer contains 512 neurons, and the third layer contains 64 neurons. The output layer uses the Softmax activation function to adapt to the classification task of 3 disease types and 3 risk levels, meeting the specific disease diagnosis needs. The He initialization method was used to initialize the parameters of the replaced fully connected layers to ensure that the parameters of the new layers are reasonably distributed and to avoid gradient vanishing or exploding in the early stage of training. The initial learning rate was set to 0.001 to balance the model convergence speed and training stability, thus completing the model framework initialization. EEG time-domain features, frequency-domain features, and correlation features between EEG signals and molecular marker expressions are selected from cross-modal joint feature vectors. Based on the model output initialized by transfer learning, a brain functional zoning atlas framework including the prefrontal, parietal, temporal, occipital, and hippocampal regions is established by combining brain anatomical structural zoning. The feature values ​​corresponding to the EEG data acquired in each adhesion-acquisition cycle are filled into the atlas framework according to brain functional zoning to form an initial personalized EEG feature atlas. The model parameters are updated using a personalized adaptive incremental training algorithm. The model parameter update formula for the personalized adaptive incremental training algorithm is as follows: ,in, For the first Model parameters after the second data collection For the first Model parameters after the second data collection For personalized forgetting factors, The novelty coefficient of the new data. For adaptive learning rate, The gradient of the cross-entropy loss function. It is the final output cross-modal joint feature vector. A map of brain function-molecular mechanisms. For diagnostic labels, For the first The effective data volume of each collection The data volume reference value is used to correct the characteristic weight and correlation strength of each brain function partition, strengthen the representation of individual-specific EEG features of the patient, and avoid the neglect of individual differences by the general model. The atlas is updated synchronously according to the number of adhesion-collection cycles, and the dynamic change trajectory of the EEG features of each partition at different time nodes is recorded to form a complete personalized EEG feature atlas, which provides accurate basis for individualized diagnosis and prescription. The cross-modal joint feature vector, personalized EEG feature atlas and brain function-molecular mechanism correlation atlas are synchronously input into the model optimized by the personalized adaptive incremental training algorithm: the model strengthens the correlation mapping between individual-specific EEG features and molecular expression rules through the feature extraction network, improves the understanding of the patient's individual condition, and outputs a disease diagnosis result that is more in line with the actual situation of the patient after classification calculation by the full connection layer. Based on the pathological mechanism correlation explanation in the diagnosis result, the EEG abnormal feature and the expression rule of the molecular marker corresponding to the brain function partition are extracted; combined with the positioning result of the brain function partition, the core parameters of the individualized neural regulation prescription are determined: the stimulation target is the brain function partition corresponding to the EEG abnormal feature, ensuring that the stimulation directly acts on the lesion-related area; the stimulation frequency matches the characteristics of the EEG abnormal wave band, improving the regulation effect of the stimulation on the abnormal EEG; the stimulation intensity is adapted according to the individual skin impedance and EEG signal amplitude to avoid discomfort caused by excessive intensity or affect the effect due to insufficient intensity; the stimulation duration is related to the expression concentration of the molecular marker, so that the intervention is matched with the molecular level change; the above parameters are integrated to form a complete individualized neural regulation prescription, which is synchronously output to the system terminal with the disease diagnosis result to provide accurate instruction support for subsequent collaborative intervention.

[0041] The collaborative intervention is executed: the adhesion state of the hydrogel electrode is detected in real time by the sensor integrated with the electrode, and the skin temperature of the electrode attachment area is synchronously monitored: the actual adhesion pressure of the hydrogel electrode is ensured to be ≥5kPa, the stable attachment of the electrode and the brain skin is ensured, and the influence of the electrode falling off on the intervention effect in the stimulation process is avoided; the skin temperature is ensured to be ≤38℃, the skin overheating caused by long-time stimulation is prevented, and the safety of the patient in the intervention process is ensured. The collaborative intervention parameter optimization algorithm is adopted, the ratio of the EEG abnormal feature improvement amount to the total intervention time and the ratio of the physiological parameter fluctuation amount to the molecular marker concentration decrease amount are comprehensively considered, the weight coefficient is balanced to balance the intervention effect and the physiological stability of the patient, and the optimal neural regulation parameter and the drug dose are determined. The mathematical expression of the collaborative intervention parameter optimization algorithm is: ; the constraint condition is kPa, , wherein, is the optimal neural regulation parameter, is the optimal stimulation frequency, is the optimal stimulation intensity, is the optimal stimulation duration, is the optimal drug dose, , a weight coefficient, a brain electrical abnormality feature improvement amount, a total intervention duration, a physiological parameter fluctuation amount, a molecular marker concentration reduction amount, an actual adhesion pressure of the hydrogel electrode, a skin temperature of the electrode attachment area, wherein the optimal neuromodulation parameters include an optimal stimulation frequency, an optimal stimulation intensity, and an optimal stimulation duration, ensuring that the electrical stimulation and drug delivery are synergistic, improving the intervention effect while reducing adverse effects on the patient's physiological state. The transcranial electrical stimulation and drug delivery operations are performed in sequence, and monitoring data is continuously collected throughout the process to monitor changes in the patient's brain electrical, physiological, molecular, and behavioral characteristics during the intervention, providing a basis for subsequent effect evaluation and parameter adjustment.

[0042] Closed-loop optimization update: at a preset time node, the multi-modal data of the patient after intervention is collected again and supplemented to the dynamic data set, enriching the data set's coverage of the post-intervention state, enabling the model to adjust its understanding of the patient's condition based on the latest data. A closed-loop effect feedback optimization algorithm is used to compare the differences between the post-intervention cross-modal joint feature vector and the pre-intervention cross-modal joint feature vector, calculate the parameter update amount based on the optimization coefficient and the synergistic effect index, adjust the model parameters, intervention parameters, and data collection frequency. The mathematical expression of the closed-loop effect feedback optimization algorithm is: wherein, is the parameter update amount, is the optimization coefficient, is the synergistic effect index, is the post-intervention cross-modal joint feature vector, is the pre-intervention cross-modal joint feature vector; the optimized model parameters improve the adaptability to post-intervention data, the adjusted intervention parameters make subsequent interventions more suitable for the patient's condition changes after intervention, and the updated collection frequency optimizes data collection efficiency according to the rhythm of condition changes, avoiding unnecessary frequent collection or data loss, forming a cycle optimization link to continuously improve diagnostic accuracy and intervention effect.

[0043] In summary, the present embodiment is around the medical data processing of Alzheimer's disease patients, through five steps to form a complete process: first, with specific parameters of semi-dry electrode to fit the parietal lobe and hippocampal region, combined with multiple devices to collect multi-dimensional data, to provide comprehensive information support for diagnosis; then through the pre-processing and cross-modal spatio-temporal attention fusion algorithm, generate joint feature vector and brain function-molecular mechanism correlation map, improve data effectiveness; then based on the transfer learning and personalized adaptive incremental training algorithm to construct the model, output accurate diagnosis and personalized neuroregulation prescription; then through the sensor monitoring and collaborative intervention parameter optimization algorithm to execute safe intervention; finally, the closed-loop feedback optimization algorithm is used to adjust the parameters in a cycle, continuously improve the diagnosis accuracy and intervention effect, and adapt to the dynamic changes of the patient's condition.

[0044] Embodiment two:

[0045] Medical data processing system embodiment for epilepsy patients.

[0046] Data acquisition module: This module is composed of semi-dry electrodes and supporting acquisition equipment. The core function is to obtain multi-dimensional dynamic data of epilepsy patients. The core conductive layer of the semi-dry electrode is a double-network interpenetrating structure composed of PVA network and PAM network in a mass ratio of 2.5:1, and an electric response characteristic enhancer accounting for 4% of the total mass of the hydrogel is added. After 3 freeze-thaw cycles, the electrode is formed. The electrode has a stretchability of 1350%, a tensile strength of 290kPa, and an electric adhesion reversible number of ≥16 times. The adhesion force is released within 28s after power-off. A micro-programmable transcranial alternating current stimulation array is integrated. The high stretchability and tensile strength enable the electrode to adapt to the small movements of the brain during seizures, tightly fit the temporal lobe and frontal lobe, and support long-term cyclic acquisition with multiple reversible adhesion characteristics, reducing the stimulation of the scalp of the patient caused by frequent electrode replacement and reducing the patient's resistance. The supporting acquisition equipment includes Ag / AgCl EEG sensor, micro piezoresistive pulse sensor, NTC thermistor temperature sensor, CdSe / ZnS core-shell structure quantum dot marker probe, near-infrared fluorescence imaging system, and wearable acceleration sensor: The Ag / AgCl EEG sensor accurately captures the abnormal discharge signals related to seizures, providing the core electrophysiological basis for diagnosis; The micro piezoresistive pulse sensor and NTC thermistor temperature sensor monitor the patient's pulse frequency and skin temperature in real time to master the patient's basic physiological state during intervention; The CdSe / ZnS core-shell structure quantum dot marker probe combined with the near-infrared fluorescence imaging system clearly obtains the expression concentration and spatial distribution data of specific molecular markers related to epilepsy in the brain, revealing the pathological changes at the molecular level; The wearable acceleration sensor conveniently records behavior characteristic data such as walking speed, limb tremor frequency, and limb movement amplitude, reflecting the patient's daily functional status. The module acquires the above multi-dimensional data in a cycle to form a dynamic data set, providing comprehensive and continuous raw data input for the subsequent module, avoiding diagnosis bias caused by a single data type, and accurately recording the time interval of each data acquisition during the acquisition process. The acquisition time and time stamp of each modality data are stored synchronously, providing the basic data required for time sequence decay coefficient calculation for the cross-modality spatio-temporal attention fusion algorithm in the subsequent data preprocessing and fusion module, ensuring the time sequence accuracy of cross-modality feature fusion, such as the medical data processing system flowchart shown in Figure 2 The five modules can be clearly understood in the medical data processing system flowchart: the data acquisition module acquires multi-modality data and transmits it to the preprocessing and fusion module, the fused data is input into the model training and prescription generation module to output diagnosis and prescription, the intervention execution module implements intervention according to the prescription, and the closed-loop optimization and update module adjusts the parameters of each module with intervention data to ensure that the system dynamically adapts to the diagnosis and treatment needs.

[0047] Data preprocessing and fusion module: As the core link of system data processing, after receiving the multi-modal data transmitted by the data acquisition module, the module first performs denoising, normalization, and spatial registration preprocessing: denoising operation removes irrelevant signals such as environmental electromagnetic interference and patient electromyographic interference, ensuring the accuracy of core data such as electroencephalogram; normalization processing unifies the dimension of different modal data, avoiding the influence of data range difference on the fairness of feature fusion; spatial registration accurately aligns each modal data with the anatomical structure determined by brain MRI, ensuring the consistency of molecular expression imaging data, electroencephalogram data, etc. in spatial position, laying a foundation for subsequent correlation analysis. Subsequently, the module defines the brain function partition of prefrontal lobe, parietal lobe, temporal lobe, occipital lobe and hippocampus based on brain anatomy, determines the spatial coordinate range of each partition combined with MRI image, extracts the expression information of molecular markers in each partition from the preprocessed molecular expression imaging data, quantifies the correlation strength between brain function and molecular mechanism, constructs and corrects the brain function-molecular mechanism correlation map, clearly presents the corresponding relationship between brain function abnormalities and molecular changes, and provides intuitive reference for model understanding of the molecular mechanism of epilepsy. Finally, the module adopts a cross-modal spatio-temporal attention fusion algorithm, which can automatically identify the key features related to epilepsy diagnosis and intervention in each modality, weaken irrelevant information, and generate a cross-modal joint feature vector through weighted splicing. The mathematical expression of the cross-modal spatio-temporal attention fusion algorithm is: wherein, is the final output cross-modal joint feature vector, corresponds to the brain electrical, molecular imaging, physiological, and behavioral four modalities, is the modal reference weight, is the modal reliability factor, is the feature vector of each modality, is the spatial correlation matrix, is the time series decay coefficient, is the acquisition time interval, integrating electro-physiological, physiological, molecular, and behavioral multi-dimensional information, avoiding the limitations of single modal data, and providing high-quality and high-dimensional feature data support for model training and prescription generation modules.

[0048] Model Training and Prescription Generation Module: This module incorporates a model initialization framework based on transfer learning. The PhysioNet EEG dataset is selected as the pre-training dataset. The base model is a ResNet-18 network containing 16 convolutional layers, 4 pooling layers, and 3 fully connected layers. The parameters of the 16 convolutional layers and 4 pooling layers in the pre-trained model are retained; these parameters contain general EEG feature extraction capabilities, reducing the computational cost of training the model from scratch and accelerating the initialization speed. Simultaneously, the broad representativeness of the public dataset improves the model's basic performance. The original fully connected layers are replaced with a 3-layer adaptive structure, and the output layer uses the Softmax activation function to adapt to classification tasks involving 3 disease types and 3 risk levels, meeting the specific needs of epilepsy diagnosis. The He initialization method is used to initialize the parameters of the new fully connected layers, ensuring a reasonable parameter distribution and avoiding gradient problems in the early stages of training. The initial learning rate is set to 0.001 to balance model convergence speed and training stability, completing the model framework initialization. After receiving the cross-modal joint feature vector and the brain function-molecular mechanism association map, the module selects EEG time-domain features, frequency-domain features, and EEG-molecular association features from the feature vector. Combined with the output of the transfer learning model, it establishes a brain function zoning map framework, filling in the EEG features collected each time according to the zoning to form an initial personalized EEG feature map. The model parameters are iteratively updated through a personalized adaptive incremental training algorithm. The model parameter update formula for the personalized adaptive incremental training algorithm is: ,in, For the first Model parameters after the second data collection For the first Model parameters after the second data collection For personalized forgetting factors, The novelty coefficient of the new data. For adaptive learning rate, The gradient of the cross-entropy loss function. It is the final output cross-modal joint feature vector. A map of brain function-molecular mechanisms. For diagnostic labels, For the first The effective data volume of each collection Using the data volume as a baseline, the feature weights and correlation strengths of each partition are adjusted to enhance the representation of individual patient-specific epilepsy EEG characteristics. The atlas is updated synchronously according to the acquisition cycle, recording feature trajectories at different time points to form a complete personalized EEG feature atlas, accurately capturing individual patient differences. The module inputs the above data into the optimized model, which strengthens individual-specific correlation mapping through a feature extraction network, improving the accuracy of epilepsy diagnosis results. Based on the pathological mechanism description in the diagnosis results, combined with brain functional partition localization, the core parameters of neuromodulation are determined, integrated to form a personalized neuromodulation prescription, and output synchronously with the diagnosis results to the collaborative intervention execution module, providing precise instructions for intervention.

[0049] Collaborative Intervention Execution Module: This module is equipped with an electrode adhesion status sensor, a transcranial electrical stimulation execution component, and a drug delivery component. Its core function is to safely and accurately execute intervention procedures. After receiving the diagnostic results and neuromodulation prescription, the module uses integrated sensors to monitor the actual adhesion pressure of the hydrogel electrodes and the skin temperature of the electrode contact area in real time. Adhesion pressure monitoring ensures stable electrode contact, preventing detachment during stimulation that could interrupt the intervention or weaken its effect. Skin temperature monitoring prevents skin burns caused by prolonged electrical stimulation, ensuring patient safety. Subsequently, the module employs a collaborative intervention parameter optimization algorithm, comprehensively considering the improvement in abnormal EEG features, total intervention duration, physiological parameter fluctuations, and decrease in molecular marker concentrations. This is combined with weighted coefficients to balance the intervention effect and the patient's physiological stability, determining the optimal neuromodulation parameters and drug dosage. The mathematical expression for the collaborative intervention parameter optimization algorithm is: The constraints are: kPa ,in, The ultimate goal is to find the optimal combination of parameters. For mathematical operators, find the independent variable that maximizes the function. ) value, To achieve optimal neural regulation parameters, For optimal stimulation frequency, For optimal stimulus intensity, For the optimal stimulus duration, To achieve the optimal drug dosage, , These are the weighting coefficients. The amount of improvement in abnormal EEG characteristics. To intervene in the total duration, This refers to the fluctuation of physiological parameters. This represents the decrease in molecular marker concentration. This represents the actual adhesion pressure of the hydrogel electrode. The skin temperature of the electrode attachment area is ensured to ensure the synergistic effect of electrical stimulation and drugs, to improve the inhibitory effect on abnormal discharge of epilepsy, and to reduce the side effects of drugs and the interference of electrical stimulation on physiological state. The module control execution component sequentially executes transcranial electrical stimulation and drug delivery, continuously collects electroencephalogram, physiological, molecular and behavioral monitoring data during the whole process, and real-time feedbacks the changes of patient state during intervention, to provide intervention effect data support for closed-loop optimization update module.

[0050] The closed-loop optimization update module is the core of the continuous optimization of the system. The module triggers the data acquisition module at the preset time node, and collects the multi-modal data of the patient after intervention again. The new data is supplemented to the dynamic data set, the coverage of the data set to the post-intervention state is enriched, so that the system can adjust the strategy based on the latest changes of the disease. The module uses a closed-loop effect feedback optimization algorithm. The cross-modal joint feature vectors after intervention and before intervention are compared, the optimization coefficient and the synergistic effect index are calculated to update the parameters. The mathematical expression of the closed-loop effect feedback optimization algorithm is: wherein, is the parameter update amount, is the optimization coefficient, is the synergistic effect index, is the cross-modal joint feature vector after intervention, is the cross-modal joint feature vector before intervention, and adjustment instructions are sent to the data acquisition module, the data preprocessing and fusion module, the model training and prescription generation module, and the synergistic intervention execution module: adjusting the data acquisition frequency to adapt the collection rhythm to the changes of the patient's disease; optimizing the model parameters to improve the adaptability and diagnostic accuracy of the model to the post-intervention data; correcting the intervention parameters to make the subsequent stimulation and drug dosage more suitable for the patient's pathological state after intervention; updating the preprocessing and fusion strategy to ensure that the data processing always matches the latest data features. Through the cyclic adjustment of the parameters of multiple modules, a closed-loop optimization link is formed to continuously improve the system's diagnostic accuracy and intervention effect, and adapt to the dynamic changes of the patient's disease.

[0051] In summary, the medical data processing system for epilepsy patients in the embodiment relies on the synergistic operation of the five modules: the data acquisition module uses special hydrogel electrodes and multiple devices to accurately collect multi-dimensional data related to the temporal lobe and frontal lobe, reducing patient discomfort; the data preprocessing and fusion module generates high-quality features and correlation graphs through preprocessing operations and cross-modal spatio-temporal attention fusion algorithms; the model training and prescription generation module outputs personalized diagnosis and prescriptions based on transfer learning models and incremental algorithms; the synergistic intervention execution module monitors adhesion and temperature, and uses optimization algorithms to achieve safe synergistic intervention; the closed-loop optimization update module supplements data at the node, adjusts the parameters of multiple modules through algorithms, forms a cyclic optimization link, and ensures that the system continuously adapts to the disease of epilepsy patients, and improves the diagnosis and treatment effect.

[0052] The above merely describes the preferred embodiments of the present application, and is not intended to limit the present application in any form. Although the present application has been disclosed with the preferred embodiments as above, it is not intended to limit the present application. Any person skilled in the art can make some changes or modifications to the above disclosed technical content to obtain equivalent embodiments with equivalent changes, as long as the changes or modifications do not deviate from the technical solution of the present application. Any modification, change, equivalent change and modification of the above embodiments made according to the technical essence of the present application still belong to the scope of the technical solution of the present application.

Claims

1. A medical data processing method for reversible adhesion of hydrogel brain electrodes, characterized in that, The specific steps of this method are as follows: S100, Multimodal Dynamic Acquisition: A semi-dry brain electrode integrating a micro-programmable transcranial AC stimulation array is attached to the target functional area of ​​the patient's brain. The core conductive layer of the semi-dry brain electrode is a conductive hydrogel with a dual-network interpenetrating structure composed of PVA network and PAM network. By applying a continuous DC voltage to the conductive hydrogel, and through the electrodes and supporting acquisition equipment, EEG signals, physiological parameters, molecular expression imaging data and behavioral feature data are cyclically acquired to form a multi-dimensional dynamic dataset. S200, Data Preprocessing and Fusion: Denoising, normalization, and spatial registration preprocessing are performed on the collected multimodal data respectively. A cross-modal spatiotemporal attention fusion algorithm is used to extract key features of each modality and weighted concatenation to generate cross-modal joint feature vectors and brain function-molecular mechanism association maps. S300, Model Training and Prescription Generation: Based on the transfer learning initialization model framework, a personalized adaptive incremental training algorithm is adopted, and the model parameters are iteratively updated using dynamic datasets to construct personalized EEG feature maps and output disease diagnosis results and personalized neuromodulation prescriptions. S400, Collaborative Intervention Execution: The adhesion status is detected by sensors integrated with electrodes, and the optimal neuromodulation parameters and drug dosage are determined by a collaborative intervention parameter optimization algorithm. Transcranial electrical stimulation and drug delivery are executed sequentially, and monitoring data are continuously collected throughout the process. S500, closed-loop optimization and update: Data after intervention is collected again at preset time nodes and added to the dynamic dataset. The closed-loop effect feedback optimization algorithm is used to calculate the parameter update amount, adjust the model parameters, intervention parameters and collection frequency, and form a cyclic optimization link.

2. The medical data processing method for reversible adhesion of hydrogel brain electrodes according to claim 1, characterized in that, In the S100 multimodal dynamic acquisition, the conductive hydrogel of the semi-dry brain electrode is formed by a double-network interpenetrating structure composed of PVA network and PAM network in a mass ratio of 2:1-3:

1. An electro-responsive property enhancer accounting for 3%-6% of the total mass of the hydrogel is added. After 3-5 freeze-thaw cycles, it is molded with an elongation ≥1200%, tensile strength ≥280kPa, and electro-adhesion reversibility ≥15 cycles. The adhesion force is released within 30 seconds after power failure. The target functional areas of the patient's brain include: the parietal lobe and hippocampus of Alzheimer's patients, and the temporal lobe and frontal lobe of epilepsy patients. The adhesion position is determined according to the lesion-associated areas located by brain MRI. The supporting acquisition equipment includes: an Ag / AgCl EEG sensor, a miniature piezoresistive pulse sensor, an NTC thermistor temperature sensor, a CdSe / ZnS core-shell quantum dot labeled probe, a near-infrared fluorescence imaging system, and a wearable accelerometer.

3. The medical data processing method for reversible adhesion of hydrogel brain electrodes according to claim 1, characterized in that, In the S100 multimodal dynamic acquisition, the acquired EEG signals include brain electrical activity signals, physiological parameters include pulse rate and skin temperature, molecular expression imaging data includes expression concentration data and spatial distribution data of specific molecular markers in the brain, and behavioral characteristic data includes walking speed, limb tremor frequency, and limb movement amplitude data.

4. The medical data processing method for reversible adhesion of hydrogel brain electrodes according to claim 1, characterized in that, In S200, during data preprocessing and fusion, the mathematical expression for the cross-modal spatiotemporal attention fusion algorithm is: ,in, It is the final output cross-modal joint feature vector. Corresponding to four modalities: electroencephalography (EEG), molecular imaging, physiological, and behavioral. For modal baseline weights, Modal reliability factor, For each modal feature vector, It is a spatial incidence matrix. This is the time-series decay coefficient. This represents the data collection time interval.

5. The medical data processing method for reversible adhesion of hydrogel brain electrodes according to claim 1, characterized in that, In S300, the specific steps for constructing a personalized EEG feature map in model training and prescription generation are as follows: screening core EEG-related features from cross-modal joint feature vectors, including EEG time-domain features, frequency-domain features, and correlation features between EEG signals and molecular marker expression. Based on the model output initialized by transfer learning, and combined with the brain anatomical structure partitioning, a brain functional partitioning atlas framework including the prefrontal lobe, parietal lobe, temporal lobe, occipital lobe, and hippocampus is established. The feature values ​​corresponding to the EEG data acquired in each adhesion-acquisition cycle are filled into the atlas framework according to the brain functional partitioning to form an initial personalized EEG feature atlas. By using the updated model parameters of the personalized adaptive incremental training algorithm, the feature weights and correlation strengths of each brain functional region are corrected to enhance the representation of individual-specific EEG features. The atlas is updated synchronously according to the number of adhesion-acquisition cycles to record the dynamic change trajectory of EEG features in each region at different time points, forming a complete personalized EEG feature atlas.

6. The medical data processing method for reversible adhesion of hydrogel brain electrodes according to claim 1, characterized in that, In S300, during model training and prescription generation, the specific content of the transfer learning initialization model framework is as follows: the pre-training dataset is selected as the PhysioNet EEG public dataset; the basic model is selected as the ResNet-18 network, which contains 16 convolutional layers, 4 pooling layers, and 3 fully connected layers, where the convolutional layers are used for feature extraction and the fully connected layers are used for classification output; the parameters of the 16 convolutional layers and 4 pooling layers trained on the public dataset in the pre-training model are retained as the initial weights for feature extraction; The original fully connected layer was replaced with a 3-layer adaptive structure. The first layer contains 1024 neurons, the second layer contains 512 neurons, and the third layer contains 64 neurons. The output layer uses the Softmax activation function to adapt to the classification task of 3 disease types and 3 risk levels. The parameters of the replaced fully connected layer were initialized using the He initialization method, and the initial learning rate was set to 0.001 to complete the transfer learning initialization of the model framework.

7. The medical data processing method for reversible adhesion of hydrogel brain electrodes according to claim 1, characterized in that, In step S300, during model training and prescription generation, the model parameter update formula for the personalized adaptive incremental training algorithm is as follows: ,in, For the first Model parameters after the second data collection For the first Model parameters after the second data collection For personalized forgetting factors, The novelty coefficient of the new data. For adaptive learning rate, The gradient of the cross-entropy loss function. It is the final output cross-modal joint feature vector. A map of brain function-molecular mechanisms. For diagnostic labels, For the first The effective data volume of each collection This serves as the baseline value for the amount of data.

8. A medical data processing method for reversible adhesion of hydrogel brain electrodes according to claim 1, characterized in that, In S400, during the execution of the collaborative intervention, the mathematical expression of the collaborative intervention parameter optimization algorithm is: The constraints are: kPa, ,in, The ultimate goal is to find the optimal combination of parameters. For mathematical operators, find the independent variable that maximizes the function. ) value, To achieve optimal neural regulation parameters, For optimal stimulation frequency, For optimal stimulus intensity, For the optimal stimulus duration, To achieve the optimal drug dosage, , These are the weighting coefficients. The amount of improvement in abnormal EEG characteristics, To intervene in the total duration, This refers to the fluctuation of physiological parameters. This represents the decrease in molecular marker concentration. This represents the actual adhesion pressure of the hydrogel electrode. The skin temperature at the electrode contact area.

9. A medical data processing method for reversible adhesion of hydrogel brain electrodes according to claim 1, characterized in that, In the S500 closed-loop optimization update, the mathematical expression of the closed-loop effect feedback optimization algorithm is: ,in, For parameter update amount, To optimize the coefficients, The synergistic effect index, To form a joint feature vector across modalities after intervention, To intervene in the joint feature vector of the cross-modal approaches.

10. A medical data processing system for reversibly adhered hydrogel brain electrodes, the system being applicable to the medical data processing method for reversibly adhered hydrogel brain electrodes as described in any one of claims 1-9, characterized in that, The system includes: Data acquisition module: Composed of semi-dry brain electrodes and supporting acquisition equipment. The conductive layer of the semi-dry brain electrodes is a conductive hydrogel with a double network interpenetrating structure composed of PVA network and PAM network, which has electroresponsive characteristics. It achieves electroadhesion to the scalp by applying a continuous DC voltage. It is used to fit the target functional area of ​​the patient's brain and cyclically collect brain electroencephalogram signals, physiological parameters, molecular expression imaging data and behavioral feature data to form a multi-dimensional dynamic dataset. The data preprocessing and fusion module is used to perform denoising, normalization, and spatial registration preprocessing on the collected multimodal data, and to generate cross-modal joint feature vectors through a cross-modal spatiotemporal attention fusion algorithm. Model training and prescription generation module: used to initialize the model framework based on transfer learning, iteratively update the model parameters through a personalized adaptive incremental training algorithm, construct a personalized EEG feature map, and output disease diagnosis results and personalized neuromodulation prescriptions; Collaborative intervention execution module: Used to detect electrode adhesion status through sensors, determine optimal neuromodulation parameters and drug dosage based on collaborative intervention parameter optimization algorithm, execute transcranial electrical stimulation and drug delivery, and continuously collect monitoring data; Closed-loop optimization and update module: It is used to collect post-intervention data at preset time nodes, supplement the dynamic dataset, and adjust the model parameters, intervention parameters and collection frequency through closed-loop effect feedback optimization algorithm to form a cyclic optimization link.

Images