A transcranial stimulation magnetic therapy closed-loop control method and system

Abstract

The application provides a transcranial stimulation magnetic therapy closed-loop control method and system, which comprises the following steps: determining a multi-modal physiological signal of a user, the multi-modal physiological signal at least comprising an electroencephalogram signal, an autonomic nervous physiological signal and a subjective feedback signal; constructing an individual dynamic baseline spectrum based on multi-state data of the user collected before treatment; pre-processing the multi-modal physiological signal collected during treatment and extracting a real-time feature vector; calculating a comprehensive deviation index of the real-time feature vector and the individual dynamic baseline spectrum, and determining a deviation type based on the comprehensive deviation index; dynamically calculating an adjustment amount of a transcranial magnetic stimulation parameter according to the deviation type, individual response characteristics and historical treatment data, and performing a stimulation operation according to the adjustment amount of the transcranial magnetic stimulation parameter, the transcranial magnetic stimulation parameter at least comprising a stimulation frequency, intensity, pulse width and stimulation duration, so as to adaptively change the transcranial magnetic stimulation parameter according to changes in neuron oscillation of the patient during treatment.

Description

Technical Field

[0001] This invention belongs to the field of neuromodulation technology, specifically relating to a closed-loop control method and system for transcranial magnetic stimulation therapy. Background Technology

[0002] Transcranial magnetic stimulation (TMS) is a painless, non-invasive, and green treatment method. The magnetic field penetrates the skin and skull non-invasively to stimulate the brain nerves. It uses pulsed magnetic fields to act on the central nervous system (mainly the brain), changing the membrane potential of cortical nerve cells, causing them to generate induced currents, and influencing brain metabolism and nerve electrical activity.

[0003] Currently, transcranial magnetic stimulation (TMS) is an open-loop treatment method. This involves using an EEG device to collect the patient's electroencephalogram (EEG) before treatment to observe neuronal oscillations. Then, appropriate TMS parameters are selected based on the EEG data for treatment. After treatment, the patient's EEG is collected again to compare the changes in neuronal oscillations before and after treatment to evaluate the treatment effect. This treatment method cannot adaptively adjust the TMS parameters according to changes in the patient's neuronal oscillations during treatment, thus failing to form a closed-loop TMS treatment. Summary of the Invention

[0004] Based on this, the present invention provides a closed-loop control method and system for transcranial magnetic stimulation therapy, which aims to adaptively change the parameters of transcranial magnetic stimulation according to the changes in neuronal oscillations during the treatment process.

[0005] A first aspect of this invention provides a closed-loop control method for transcranial magnetic stimulation therapy, the method comprising:

[0006] The user's multimodal physiological signals are determined, including at least electroencephalogram (EEG) signals, autonomic nervous system physiological signals, and subjective feedback signals.

[0007] An individual dynamic baseline spectrum is constructed based on multi-state data of users collected before treatment. The multi-state data includes multimodal physiological signals under resting state, task state, and simulated stimulus state.

[0008] The multimodal physiological signals collected during the treatment process are preprocessed, and real-time feature vectors are extracted.

[0009] Calculate the comprehensive deviation index between the real-time feature vector and the individual dynamic baseline spectrum, and determine the deviation type based on the comprehensive deviation index;

[0010] The adjustment amount of transcranial magnetic stimulation parameters is dynamically calculated based on the deviation type, individual response characteristics and historical treatment data, and stimulation operation is performed based on the adjustment amount of transcranial magnetic stimulation parameters. The transcranial magnetic stimulation parameters include at least stimulation frequency, intensity, pulse width and stimulation duration.

[0011] Furthermore, the transcranial magnetic stimulation closed-loop control method also includes:

[0012] A predictive model is trained based on real-time data during the treatment process to anticipate subsequent state deviations and adjust stimulation parameters in advance.

[0013] Furthermore, the step of constructing an individual dynamic baseline spectrum based on user multi-state data collected before treatment includes:

[0014] Based on the multimodal physiological signals collected before treatment under resting, task, and simulated stimulus conditions, EEG features and autonomic nervous system physiological features are extracted, and an individual baseline feature matrix is ​​formed. The EEG features include at least the relative power of each frequency band, the rate of change of each frequency band, and the frequency band coherence. The autonomic nervous system physiological features include at least the high-frequency component of heart rate variability and the average amplitude of skin conductance.

[0015] The features in the individual baseline feature matrix are labeled, and a similarity threshold is set to obtain the individual dynamic baseline spectrum, where the labeled values ​​are health reference values.

[0016] Furthermore, the step of calculating the comprehensive deviation index between the real-time feature vector and the individual dynamic baseline spectrum, and determining the deviation type based on the comprehensive deviation index, includes:

[0017] The relative deviations between the real-time feature vectors and the corresponding health reference values ​​in the individual dynamic baseline spectrum are calculated respectively, and weights are assigned to each deviation value according to the disease type. The comprehensive deviation index is obtained by weighted summation.

[0018] The deviation levels are classified according to the comprehensive deviation index, including mild deviation, moderate deviation and severe deviation, and the deviation type is determined by combining the subjective feedback signal. The deviation type includes at least insufficient brain region activation, excessive patient tension and mixed deviation.

[0019] Furthermore, the step of dynamically calculating the adjustment amount of transcranial magnetic stimulation parameters based on the deviation type, individual response characteristics, and historical treatment data, and performing stimulation operations based on the adjustment amount of transcranial magnetic stimulation parameters, includes:

[0020] For the target brain region, the stimulation frequency, intensity, and pulse width are adjusted sequentially with preset amplitudes. After each adjustment, the EEG signal and autonomic nervous physiological signal are collected at the first preset time, and the parameter change and feature improvement of each parameter dimension are recorded.

[0021] Calculate the corresponding response sensitivity based on the parameter change and feature improvement for each parameter dimension;

[0022] Record the three data points about the scenario, adjustment, and effect of each treatment to establish an individual treatment database. The scenario data includes the comprehensive deviation index before adjustment, the physiological state at that time, and the treatment stage. The adjustment data is the parameters of this adjustment, including the change in stimulation frequency, the change in intensity, and the change in pulse width. The effect data includes the decrease in the comprehensive deviation index at the second preset time after adjustment, as well as the patient's subjective feedback.

[0023] The data in the individual treatment database is labeled to obtain effective adjustment cases. Each time an adjustment is needed, the similarity between the current scenario and the effective adjustment cases in the individual treatment database is calculated.

[0024] Determine whether the scene similarity is greater than a threshold;

[0025] If so, then obtain valid adjustment cases with scene similarity greater than the threshold, extract the parameter adjustment direction and adjustment amount ratio, and determine the historical weight;

[0026] The adjustment amount of the transcranial magnetic stimulation parameters is calculated based on the baseline adjustment amount determined by the deviation type, the response sensitivity, and the historical weights.

[0027] Furthermore, the steps of obtaining valid adjustment cases with scene similarity greater than a threshold, extracting parameters such as adjustment direction and adjustment amount ratio, and determining historical weights include:

[0028] Calculate the basic contribution value of each parameter in valid adjustment cases where the scene similarity is greater than the threshold;

[0029] The basic contribution value is weighted by the scene similarity to obtain the weighted contribution value;

[0030] The weighted contribution values ​​of all similar cases are aggregated, and the contribution ratio of the corresponding parameters is calculated to obtain the historical weights.

[0031] Furthermore, in the step of training a prediction model based on real-time data during the treatment process, predicting subsequent state deviations and adjusting stimulation parameters in advance, an LSTM model is trained based on the user's baseline data before treatment and treatment data of patients with similar symptoms. The input of the LSTM model is a real-time feature sequence, and the output is the feature prediction value at a third preset time in the future.

[0032] A second aspect of the present invention provides a transcranial magnetic stimulation (TMS) closed-loop control system for implementing the TMS closed-loop control method provided in the first aspect of the present invention, the system comprising:

[0033] The determination module is used to determine the user's multimodal physiological signals, which include at least electroencephalogram (EEG) signals, autonomic nervous system physiological signals, and subjective feedback signals.

[0034] The module is used to construct an individual dynamic baseline spectrum based on multi-state data of the user collected before treatment. The multi-state data includes multimodal physiological signals under resting state, task state, and simulated stimulus state.

[0035] The extraction module is used to preprocess the multimodal physiological signals collected during the treatment process and extract real-time feature vectors.

[0036] The first calculation module is used to calculate the comprehensive deviation index between the real-time feature vector and the individual dynamic baseline spectrum, and to determine the deviation type based on the comprehensive deviation index;

[0037] The second calculation module is used to dynamically calculate the adjustment amount of the transcranial magnetic stimulation parameters based on the deviation type, individual response characteristics and historical treatment data, and to perform stimulation operation based on the adjustment amount of the transcranial magnetic stimulation parameters. The transcranial magnetic stimulation parameters include at least stimulation frequency, intensity, pulse width and stimulation duration.

[0038] A third aspect of the present invention provides a computer-readable storage medium, comprising:

[0039] The readable storage medium stores one or more programs that, when executed by a processor, implement the transcranial magnetic stimulation closed-loop control method as described in the first aspect.

[0040] A fourth aspect of the present invention provides an electronic device, the electronic device including a memory and a processor, wherein:

[0041] The memory is used to store computer programs;

[0042] When the processor executes the computer program stored in the memory, it implements the transcranial magnetic stimulation closed-loop control method as described in the first aspect.

[0043] This invention provides a closed-loop control method and system for transcranial magnetic stimulation (TMS) therapy. The method involves determining the user's multimodal physiological signals, including at least electroencephalogram (EEG) signals, autonomic nervous system physiological signals, and subjective feedback signals. An individual dynamic baseline spectrum is constructed based on multi-state data collected before treatment, including multimodal physiological signals under resting, task, and simulated stimulation states. The multimodal physiological signals collected during treatment are preprocessed, and real-time feature vectors are extracted. A comprehensive deviation index between the real-time feature vectors and the individual dynamic baseline spectrum is calculated, and the deviation type is determined based on the comprehensive deviation index. The adjustment amount of the TMS parameters is dynamically calculated based on the deviation type, individual response characteristics, and historical treatment data. Stimulation is then performed based on the adjustment amount of the TMS parameters, which include at least stimulation frequency, intensity, pulse width, and stimulation duration. Attached Figure Description

[0044] Figure 1 This is a flowchart illustrating the implementation of a closed-loop control method for transcranial magnetic stimulation therapy provided in Embodiment 1 of the present invention.

[0045] Figure 2 This is a structural block diagram of a transcranial magnetic stimulation closed-loop control system provided in Embodiment 2 of the present invention;

[0046] Figure 3 This is a structural block diagram of an electronic device provided in Embodiment 3 of the present invention. Detailed Implementation

[0047] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0048] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0049] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0050] Example 1

[0051] Embodiment 1 of this invention provides a closed-loop control method for transcranial magnetic stimulation therapy. Please refer to [link to documentation]. Figure 1 This is a flowchart of a closed-loop control method for transcranial magnetic stimulation therapy, specifically including steps S01 to S05.

[0052] Step S01: Determine the user's multimodal physiological signals, which include at least electroencephalogram (EEG) signals, autonomic nervous system physiological signals, and subjective feedback signals.

[0053] In this embodiment of the invention, a high-precision EEG device is used to collect EEG signals. At the same time, a photoplethysmography sensor and an electroskin response sensor are used to collect heart rate variability and skin conductivity, respectively. In addition, subjective feedback signals are received through an interactive terminal. These subjective feedback signals include at least relaxation, tension, and no obvious feeling.

[0054] Step S02: Construct an individual dynamic baseline spectrum based on the user's multi-state data collected before treatment. The multi-state data includes multimodal physiological signals under resting state, task state, and simulated stimulus state.

[0055] Specifically, the resting state refers to the basic EEG and physiological signals collected when the user closes their eyes and relaxes; the task state refers to the data collected when the user completes a simple cognitive task (such as the number Stroop task, which measures attention control ability) and the brain function activation data; the simulated stimulus state refers to the collection of EEG and physiological stress response data after stimulation by using low-intensity (50% of the motor threshold) TMS stimulation of the target brain region.

[0056] It's important to note that neuronal oscillations are divided into several main bands based on their frequency, each associated with specific brain states and cognitive functions. Delta (δ) waves (0.5-4 Hz) are primarily associated with deep sleep, coma, and the healing process. Theta (θ) waves (4-8 Hz) are primarily associated with drowsiness, meditation, light sleep, memory encoding and retrieval (especially in the hippocampus), and spatial navigation. Alpha (α) waves (8-13 Hz) are primarily associated with wakefulness and rest, relaxation with eyes closed, inhibiting irrelevant sensory input, and helping to "shield" interference. Beta (β) waves (13-30 Hz) are primarily associated with wakefulness, active cognitive activity, focus, problem-solving, and motor control. Gamma (γ) waves (30-100+ Hz) are primarily associated with wakefulness, active cognitive activity, focus, problem-solving, and motor control. For example, in Alzheimer's disease, gamma waves are observed; transcranial magnetic stimulation (TMS) therapy makes gamma waves more regular.

[0057] More specifically, based on the multimodal physiological signals collected before treatment under resting, task, and simulated stimulus conditions, EEG features and autonomic nervous system physiological features are extracted, and an individual baseline feature matrix is ​​formed. The EEG features include at least the relative power of each frequency band, the rate of change of each frequency band, and the frequency band coherence. Taking the γ (Gamma) wave as an example, the relative power of the frequency band refers to the ratio of the γ wave to the total power of the entire frequency band, the rate of change of the frequency band refers to the change of the γ wave within a preset time, and the frequency band coherence refers to the power ratio of the α wave to the β wave, reflecting the attention state. The autonomic nervous system physiological features include at least the high-frequency component of heart rate variability and the average amplitude of skin conductance. The high-frequency component of heart rate variability is used to reflect parasympathetic nerve activity, and the average amplitude of skin conductance reflects the excitability of the sympathetic nerve.

[0058] The features in the individual baseline feature matrix are labeled, and a similarity threshold is set to obtain the individual dynamic baseline spectrum. The labeled features are health reference values. In this embodiment of the invention, the similarity threshold is 85%, that is, when the similarity between the real-time features and the baseline features during treatment is ≥85%, it is determined to be close to a healthy state.

[0059] Step S03: Preprocess the multimodal physiological signals collected during the treatment process and extract real-time feature vectors.

[0060] Specifically, independent component analysis (ICA) was used to remove interference from electrooculography (EOG) and electromyography (EMG), and wavelet transform was used to separate the alpha, beta, and gamma frequency bands; moving average filtering (with a 5-second window) was used to smooth the noise in the physiological signals. Furthermore, EEG features and autonomic nervous system physiological features consistent with the baseline spectrum were extracted in real time to form a real-time feature vector.

[0061] Step S04: Calculate the comprehensive deviation index between the real-time feature vector and the individual dynamic baseline spectrum, and determine the deviation type based on the comprehensive deviation index.

[0062] It should be noted that the relative deviations between the real-time feature vector and the corresponding health reference values ​​in the individual dynamic baseline spectrum are calculated separately, and weights are assigned to each deviation value according to the disease type. The comprehensive deviation index is obtained by weighted summation. It can be understood that the relative deviation (D) = |real-time feature value - baseline feature value| / baseline feature value × 100%, and the comprehensive deviation index DI = (w1 × D1) + (w2 × D2) + ... + (w n ×D n ), where D n w represents the relative deviation of the nth feature. n The weight of the disease type corresponding to the nth feature is DI, which ranges from 0 to 100%. The higher the DI, the more serious the deviation from the healthy state. For example, the weight of Gamma deviation is higher in depression and the weight of physiological deviation is higher in anxiety.

[0063] The deviation levels are classified according to the comprehensive deviation index, including mild deviation, moderate deviation, and severe deviation. The deviation type is determined in conjunction with the subjective feedback signal. The deviation type includes at least insufficient brain region activation, excessive patient tension, and mixed deviation. In this embodiment of the invention, a comprehensive deviation index of 10-30% is considered mild deviation, with a base value range of ±1~2Hz for frequency adjustment as an example; a comprehensive deviation index of 31-60% is considered moderate deviation, with a base value range of ±2~4Hz for frequency adjustment as an example; and a comprehensive deviation index >60% is considered severe deviation, with a base value range of ±4~6Hz for frequency adjustment as an example. Furthermore, three buttons ("Relax", "Tense", and "No Obvious Sensation") are set on the treatment interface of the interactive terminal. The system converts subjective feedback into weights (e.g., "Tense" weight 0.95, "Relax" weight 1.05), and combines them with the objective feature similarity to form the final judgment result. For example, after calculating the comprehensive deviation index, the comprehensive deviation index is converted into similarity according to the preset mapping relationship. If the similarity value is 82% at this time and the subjective feedback is "Relax", then the final similarity value is 82% × 1.05 = 86.1%, which is greater than the similarity threshold of 85%. Therefore, it is finally judged as close to good and no adjustment is made for the time being.

[0064] Step S05: Dynamically calculate the adjustment amount of transcranial magnetic stimulation parameters based on the deviation type, individual response characteristics and historical treatment data, and perform stimulation operation based on the adjustment amount of transcranial magnetic stimulation parameters. The transcranial magnetic stimulation parameters include at least stimulation frequency, intensity, pulse width and stimulation duration.

[0065] Specifically, for the target brain region, the stimulation frequency, intensity, and pulse width are adjusted sequentially with preset amplitudes. After each adjustment, EEG signals and autonomic nervous physiological signals are collected at the first preset time, and the parameter changes and feature improvements of each parameter dimension are recorded. For example, for the stimulation frequency, 3Hz, 5Hz, and 7Hz are tried sequentially (each adjustment +2Hz), and the decrease in Gamma deviation value after each frequency change is recorded (e.g., from 60% to 50%, an improvement of 10%).

[0066] Based on the parameter change and feature improvement for each parameter dimension, the corresponding response sensitivity is calculated, and then the sensitivity coefficient of each parameter is normalized. The sensitivity coefficient (S) = feature improvement / parameter adjustment. For example, if the frequency changes from 3Hz to 5Hz (adjustment + 2Hz) and the Gamma deviation changes from 60% to 45% (improvement 15%), then the frequency sensitivity coefficient = 15% / 2Hz = 7.5% / Hz.

[0067] Record the three-dimensional data of each treatment regarding the scenario, adjustment, and effect to establish an individual treatment database. Scenario data includes the comprehensive deviation index before adjustment, the physiological state at that time (such as heart rate, skin conductance), and the treatment stage (such as week 1 / week 4). Adjustment data consists of the parameters of this adjustment, including the amount of change in stimulation frequency, intensity, and pulse width. Effect data includes the decrease in the comprehensive deviation index at the second preset time after adjustment, as well as the patient's subjective feedback.

[0068] The data in the individual treatment database are labeled to obtain effective adjustment cases. For example, if the DI decreases by ≥10% after adjustment and the subjective feedback is positive, it is marked as an effective adjustment case. Each time an adjustment is needed, the scene similarity between the current scene and the effective adjustment cases in the individual treatment database is calculated, where cosine similarity is used for calculation, with a range of 0-1.

[0069] Determine whether the scene similarity is greater than a threshold;

[0070] If so, then obtain the effective adjustment cases where the scene similarity is greater than the threshold, extract the parameter adjustment direction and adjustment amount ratio, and determine the historical weight. Specifically, calculate the basic contribution value of each parameter in the effective adjustment cases where the scene similarity is greater than the threshold, where the basic contribution value = parameter adjustment amount × individual sensitivity coefficient of the parameter.

[0071] The basic contribution value is weighted by the scene similarity to obtain the weighted contribution value, where the weighted contribution value = basic contribution value × similarity between the case and the current scene;

[0072] The weighted contribution values ​​of all similar cases are summarized, and the contribution ratio of the corresponding parameter is calculated to obtain the historical weight. The historical weight is calculated as: historical weight = total weighted contribution value of a certain parameter ÷ sum of total weighted contribution values ​​of all parameters. For example, it is assumed that there are two cases, Case 1 and Case 2, whose similarity to the current scene is greater than the threshold of 0.8. Specifically, Case 1 has a similarity of 0.9, the stimulus frequency is adjusted to +3Hz, the intensity is adjusted to +2%MT, and the stimulus frequency after normalization of the individual sensitivity coefficient is 0.6 and the intensity is 0.3. Case 2 has a similarity of 0.85, the stimulus frequency is adjusted to +2Hz, the intensity is adjusted to +5%MT, and the stimulus frequency after normalization of the individual sensitivity coefficient is 0.6 and the intensity is 0.3.

[0073] It should be noted that in Case 1: the frequency base contribution value = 3Hz × 0.6 (frequency sensitivity coefficient) = 1.8, and the intensity base contribution value = 2%MT × 0.3 (intensity sensitivity coefficient) = 0.6; in Case 2: the frequency base contribution value = 2Hz × 0.6 = 1.2, and the intensity base contribution value = 5%MT × 0.3 = 1.5.

[0074] In Case 1: Frequency-weighted contribution value = 1.8 × 0.9 (similarity) = 1.62, Intensity-weighted contribution value = 0.6 × 0.9 = 0.54; In Case 2: Frequency-weighted contribution value = 1.2 × 0.85 (similarity) = 1.02, Intensity-weighted contribution value = 1.5 × 0.85 = 1.275;

[0075] Summarizing the total weighted contribution values: Total weighted contribution value of frequency = 1.62 (Case 1) + 1.02 (Case 2) = 2.64, Total weighted contribution value of intensity = 0.54 (Case 1) + 1.275 (Case 2) = 1.815, Sum of total weighted contribution values ​​of all parameters = 2.64 + 1.815 = 4.455; Calculating historical weights: Historical weight of frequency = 2.64 ÷ 4.455 ≈ 0.59 (approximately 0.6), Historical weight of intensity = 1.815 ÷ 4.455 ≈ 0.41 (approximately 0.4), that is, stimulus frequency weight ≈ 0.6, intensity ≈ 0.4;

[0076] Based on the baseline adjustment amount determined by the deviation type, the response sensitivity, and the historical weight, the adjustment amount of the transcranial magnetic stimulation parameters is calculated. The baseline adjustment amount, response sensitivity, and historical weight are multiplied to obtain the final adjustment amount. For example, taking stimulation frequency as an example, with DI = 55% (moderate deviation), the baseline frequency range is +2 to 4 Hz. The midpoint +3 Hz is taken as the baseline adjustment amount, multiplied by the individual sensitivity coefficient weight. If the normalized frequency sensitivity coefficient is 0.6, then the adjustment amount = 3 Hz × 0.6 = 1.8 Hz (preliminary value). Multiplied by the historical weight, if the historical weight of frequency adjustment in a similar case is 0.6, then the final frequency adjustment amount = 1.8 Hz × 0.6 ≈ 1.08 Hz, rounded to +1 Hz.

[0077] In other embodiments of the present invention, in order to predict and optimize and avoid deviations in advance and improve the timeliness of adjustments, a prediction model is trained based on real-time data during the treatment process to predict subsequent state deviations and adjust stimulation parameters in advance. Specifically, an LSTM (Long Short-Term Memory) model is trained based on the user's baseline data before treatment and the treatment data of patients with similar diseases. The input of the LSTM model is a real-time feature sequence, and the output is the feature prediction value at a third preset time in the future. For example, the LSTM model is called once every 3 minutes to make a prediction. If the similarity between the predicted value and the baseline spectrum is <80% (i.e., the prediction deviation will increase), a small adjustment of parameters (such as frequency ±1Hz, intensity ±3%MT) is made in advance, rather than making a large adjustment after the deviation occurs.

[0078] In summary, the present invention proposes a closed-loop control method for transcranial magnetic stimulation (TMS) therapy. This method determines the user's multimodal physiological signals, which include at least electroencephalogram (EEG) signals, autonomic nervous system physiological signals, and subjective feedback signals. It constructs an individual dynamic baseline spectrum based on multi-state data collected before treatment, including multimodal physiological signals under resting, task, and simulated stimulation states. The multimodal physiological signals collected during treatment are preprocessed, and real-time feature vectors are extracted. A comprehensive deviation index between the real-time feature vectors and the individual dynamic baseline spectrum is calculated, and the deviation type is determined based on the comprehensive deviation index. The adjustment amount of the TMS parameters is dynamically calculated based on the deviation type, individual response characteristics, and historical treatment data. Stimulation is then performed based on the adjustment amount of the TMS parameters, which include at least stimulation frequency, intensity, pulse width, and stimulation duration. This allows for adaptive changes in the TMS parameters during treatment based on changes in the patient's neuronal oscillations.

[0079] Example 2

[0080] Embodiment 2 of the present invention provides a transcranial magnetic stimulation closed-loop control system 200. Please refer to [link / reference]. Figure 2 This is a structural block diagram of a transcranial magnetic stimulation (TMS) closed-loop control system 200, which includes:

[0081] The determination module 21 is used to determine the user's multimodal physiological signals, which include at least electroencephalogram (EEG) signals, autonomic nervous system physiological signals, and subjective feedback signals.

[0082] Module 22 is used to construct an individual dynamic baseline spectrum based on user multi-state data collected before treatment, wherein the multi-state data includes multimodal physiological signals under resting state, task state and simulated stimulus state;

[0083] Extraction module 23 is used to preprocess the multimodal physiological signals collected during the treatment process and extract real-time feature vectors;

[0084] The first calculation module 24 is used to calculate the comprehensive deviation index between the real-time feature vector and the individual dynamic baseline spectrum, and to determine the deviation type based on the comprehensive deviation index;

[0085] The second calculation module 25 is used to dynamically calculate the adjustment amount of the transcranial magnetic stimulation parameters based on the deviation type, individual response characteristics and historical treatment data, and to perform stimulation operation based on the adjustment amount of the transcranial magnetic stimulation parameters. The transcranial magnetic stimulation parameters include at least stimulation frequency, intensity, pulse width and stimulation duration.

[0086] Furthermore, in other embodiments of the present invention, the transcranial magnetic stimulation closed-loop control system 200 further includes:

[0087] The prediction module is used to train a prediction model based on real-time data during the treatment process, predict subsequent state deviations and adjust stimulation parameters in advance. Based on the user's baseline data before treatment and the treatment data of patients with similar diseases, the LSTM model is trained. The input of the LSTM model is a real-time feature sequence, and the output is the feature prediction value at a third preset time in the future.

[0088] Furthermore, in some other embodiments of the present invention, the building module 22 includes:

[0089] The extraction unit is used to extract EEG features and autonomic nervous system physiological features based on multimodal physiological signals collected before treatment under resting, task, and simulated stimulation states, and to form an individual baseline feature matrix. The EEG features include at least the relative power of each frequency band, the rate of change of each frequency band, and the frequency band coherence. The autonomic nervous system physiological features include at least the high-frequency component of heart rate variability and the average amplitude of skin conductance.

[0090] The first annotation unit is used to annotate the features in the individual baseline feature matrix and set a similarity threshold to obtain the individual dynamic baseline spectrum, wherein the annotation is a health reference value.

[0091] Furthermore, in some other embodiments of the present invention, the first computing module 24 includes:

[0092] The first calculation unit is used to calculate the relative deviation between the real-time feature vector and the corresponding health reference value in the individual dynamic baseline spectrum, and to assign weights to each deviation value according to the disease type, and to obtain the comprehensive deviation index by weighted summation.

[0093] The classification unit is used to classify the deviation level according to the comprehensive deviation index, including mild deviation, moderate deviation and severe deviation, and to determine the deviation type in combination with the subjective feedback signal. The deviation type includes at least insufficient brain region activation, excessive patient tension and mixed deviation.

[0094] Furthermore, in some other embodiments of the present invention, the second computing module 25 includes:

[0095] The acquisition unit is used to sequentially adjust the stimulation frequency, intensity, and pulse width of the target brain region by preset amplitude. After each adjustment, it acquires the EEG signal and autonomic nervous physiological signal for the first preset time and records the parameter change and feature improvement of each parameter dimension.

[0096] The second calculation unit is used to calculate the corresponding response sensitivity based on the parameter change and feature improvement for each parameter dimension.

[0097] The database establishment unit is used to record three-dimensional data about the scene, adjustment, and effect of each treatment to establish an individual treatment database. The scene data includes the comprehensive deviation index before adjustment, the physiological state at that time, and the treatment stage. The adjustment data consists of the parameters of this adjustment, including the change in stimulation frequency, the change in intensity, and the change in pulse width. The effect data includes the decrease in the comprehensive deviation index at the second preset time after adjustment, as well as the patient's subjective feedback.

[0098] The second annotation unit is used to annotate the data in the individual treatment database to obtain effective adjustment cases. Each time an adjustment is needed, the similarity between the current scene and the effective adjustment cases in the individual treatment database is calculated.

[0099] A judgment unit is used to determine whether the scene similarity is greater than a threshold.

[0100] The acquisition unit is used to acquire valid adjustment cases where the scene similarity is greater than the threshold when it is determined that the scene similarity is greater than the threshold, and extract parameters such as adjustment direction and adjustment amount ratio to determine historical weights.

[0101] The third calculation unit is used to calculate the adjustment amount of the transcranial magnetic stimulation parameters based on the baseline adjustment amount determined by the deviation type, the response sensitivity, and the historical weight.

[0102] Furthermore, in some other embodiments of the present invention, the acquiring unit includes:

[0103] The first calculation subunit is used to calculate the basic contribution value of each parameter in the effective adjustment cases where the scene similarity is greater than the threshold.

[0104] The second calculation subunit is used to weight the basic contribution value with the scene similarity to obtain a weighted contribution value;

[0105] The third calculation subunit is used to summarize the weighted contribution values ​​of all similar cases, calculate the contribution ratio of the corresponding parameters, and obtain the historical weights.

[0106] Example 3

[0107] Embodiment 3 of the present invention proposes an electronic device, please refer to [link / reference]. Figure 3 This is a structural block diagram of an electronic device, including a memory 20, a processor 10, and a computer program 30 stored in the memory and executable on the processor. When the processor 10 executes the computer program 30, it implements the transcranial magnetic stimulation closed-loop control method as described above.

[0108] In some embodiments, the processor 10 may be a central processing unit (CPU), controller, microcontroller, microprocessor or other data processing chip, used to run program code stored in memory 20 or process data, such as executing access restriction programs.

[0109] The memory 20 includes at least one type of readable storage medium, such as flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 20 can be an internal storage unit of an electronic device, such as the hard disk of the electronic device. In other embodiments, the memory 20 can also be an external storage device of the electronic device, such as a plug-in hard disk, SmartMediaCard (SMC), SecureDigital (SD) card, FlashCard, etc., equipped on the electronic device. Furthermore, the memory 20 can include both internal and external storage units of the electronic device. The memory 20 can be used not only to store application software and various types of data of the electronic device, but also to temporarily store data that has been output or will be output.

[0110] This invention also proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the transcranial magnetic stimulation closed-loop control method as described above.

[0111] Those skilled in the art will understand that the logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can mean any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0112] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0113] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0114] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0115] The above embodiments merely illustrate several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this patent should be determined by the appended claims.

Claims

1. A transcranial stimulation magnetic therapy closed-loop control system, characterized in that, The system comprises: A determination module for determining a multi-modal physiological signal of a user, the multi-modal physiological signal comprising at least an electroencephalogram signal, an autonomic nervous physiological signal, and a subjective feedback signal; A construction module for constructing an individual dynamic baseline spectrum based on multi-state data of the user collected before treatment, the multi-state data comprising multi-modal physiological signals in resting state, task state, and simulated stimulation state; An extraction module for pre-processing the multi-modal physiological signal collected during treatment and extracting a real-time feature vector; A first calculation module for calculating a comprehensive deviation index of the real-time feature vector and the individual dynamic baseline spectrum, and determining a deviation type based on the comprehensive deviation index; A second calculation module for dynamically calculating an adjustment amount of transcranial magnetic stimulation parameters according to the deviation type, individual response characteristics, and historical treatment data, and performing a stimulation operation according to the adjustment amount of transcranial magnetic stimulation parameters, the transcranial magnetic stimulation parameters comprising at least stimulation frequency, intensity, pulse width, and stimulation duration.

2. The transcranial stimulation magnetic therapy closed-loop control system of claim 1, wherein, The transcranial stimulation magnetic therapy closed-loop control system further comprises: A pre-judgment module for training a prediction model based on real-time data during treatment, pre-judging subsequent state deviation, and adjusting stimulation parameters in advance.

3. The transcranial stimulation magnetic therapy closed-loop control system of claim 2, wherein, The step of constructing an individual dynamic baseline spectrum based on multi-state data of the user collected before treatment comprises: Extracting electroencephalogram features and autonomic nervous physiological features from multi-modal physiological signals in resting state, task state, and simulated stimulation state collected before treatment, and forming an individual baseline feature matrix, the electroencephalogram features comprising at least relative power of each frequency band, change rate of each frequency band, and frequency band coherence, and the autonomic nervous physiological features comprising at least heart rate variability high-frequency component and skin electricity average amplitude; Labeling features in the individual baseline feature matrix and setting a similarity threshold to obtain the individual dynamic baseline spectrum, wherein the labeling is a healthy reference value.

4. The transcranial stimulation magnetic therapy closed-loop control system of claim 3, wherein, The step of calculating a comprehensive deviation index of the real-time feature vector and the individual dynamic baseline spectrum, and determining a deviation type based on the comprehensive deviation index comprises: Calculating the relative deviation of the real-time feature vector and the corresponding healthy reference value in the individual dynamic baseline spectrum respectively, assigning weights to each deviation value according to the disease type, and calculating the comprehensive deviation index by weighted summation; Dividing the deviation level according to the comprehensive deviation index, including mild deviation, moderate deviation, and severe deviation, and determining the deviation type in combination with the subjective feedback signal, the deviation type comprising at least insufficient activation of brain region, excessive tension of patient, and mixed deviation.

5. The transcranial stimulation magnetic therapy closed-loop control system of claim 4, wherein, The step of dynamically calculating an adjustment amount of transcranial magnetic stimulation parameters according to the deviation type, individual response characteristics, and historical treatment data, and performing a stimulation operation according to the adjustment amount of transcranial magnetic stimulation parameters comprises: For a target brain region, adjusting the stimulation frequency, intensity, and pulse width in a preset amplitude in sequence, collecting electroencephalogram signals and autonomic nervous physiological signals for a first preset time after each adjustment, and recording the parameter change amount and feature improvement amount of each parameter dimension; Calculating the corresponding response sensitivity according to the parameter change amount and the feature improvement amount of each parameter dimension; Record the triad data about scene, adjustment, and effect of each treatment to establish an individual treatment database, the scene data includes the comprehensive deviation index before adjustment, the physiological state at that time, and the treatment stage, the adjustment data is the parameter of this adjustment, including the change amount of stimulation frequency, intensity, and pulse width, and the effect data includes the decline amplitude of the comprehensive deviation index after adjustment for the second preset time, and the patient's subjective feedback; Annotate the data in the individual treatment database to obtain effective adjustment cases, and calculate the scene similarity between the current scene and the scene of the effective adjustment cases in the individual treatment database each time adjustment is needed; Determine whether the scene similarity is greater than a threshold value; If yes, obtain the effective adjustment cases with the scene similarity greater than the threshold value, extract the parameter adjustment direction and adjustment amount proportion, and determine the historical weight; Calculate the adjustment amount of the transcranial magnetic stimulation parameter according to the basic adjustment amount determined according to the deviation type, the response sensitivity, and the historical weight.

6. The transcranial stimulation magnetic therapy closed-loop control system of claim 5, wherein, The step of obtaining the effective adjustment cases with the scene similarity greater than the threshold value, extracting the parameter adjustment direction and adjustment amount proportion, and determining the historical weight comprises: Calculate the basic contribution value of each parameter in the effective adjustment cases with the scene similarity greater than the threshold value; Weight the scene similarity and the basic contribution value to obtain a weighted contribution value; Summarize the weighted contribution values of all similar cases to calculate the contribution proportion of the corresponding parameter to obtain the historical weight.

7. The transcranial stimulation magnetic therapy closed-loop control system of claim 6, wherein, In the step of training a prediction model based on real-time data during the treatment process to predict the subsequent state deviation and adjust the stimulation parameter in advance, an LSTM model is trained according to the baseline period data of the user before treatment and the treatment data of patients with the same disease, the input of the LSTM model is a real-time feature sequence, and the output is a feature prediction value in the future third preset time.

8. A computer-readable storage medium, characterized in that, Comprise: The readable storage medium stores one or more programs, which are executed by the processor to apply to the transcranial magnetic stimulation closed-loop control system in any one of claims 1-7.

9. An electronic device, comprising: The electronic device comprises a memory and a processor, wherein: The memory is used to store computer programs; The processor is used to execute the computer programs stored on the memory to apply to the transcranial magnetic stimulation closed-loop control system in any one of claims 1-7. The electronic device comprises a memory and a processor, wherein: The memory is used to store computer programs; The processor is used to execute the computer programs stored on the memory to apply to the transcranial magnetic stimulation closed-loop control system in any one of claims 1-7.

Images