Feedback training method and apparatus, electronic device, and storage medium

Abstract

The application provides a feedback training method and device, electronic equipment and a storage medium. The method comprises: acquiring an electroencephalogram signal and a physiological signal of a subject, the physiological signal being used to reflect a physiological state of the body, and the electroencephalogram signal and the physiological signal having the same timestamp; calculating an interaction index between the electroencephalogram signal and the physiological signal, the interaction index being used to represent a coupling degree between the electroencephalogram signal and the physiological signal of the subject, and the higher the value of the interaction index, the higher the coupling degree between the electroencephalogram signal and the physiological signal; mapping the interaction index to a feedback parameter, and presenting the feedback parameter to the subject to guide the subject to self-regulate the physiological state according to the feedback parameter. The feedback training effect and stability in different subjects and different application scenarios can be improved.

Description

Technical Field

[0001] This application relates to the field of medical rehabilitation technology, and in particular to a feedback training method, device, electronic device and storage medium. Background Technology

[0002] Neurofeedback and biofeedback technologies are important techniques in the field of non-invasive functional regulation and rehabilitation training. However, existing systems still belong to a single-modal feedback paradigm, typically focusing on either central nervous system activity or peripheral autonomic physiological activity, each forming a relatively independent feedback system. This leads to problems such as significant differences in training effects and insufficient stability in existing feedback training systems across different subjects and application scenarios. Summary of the Invention

[0003] This application provides a feedback training method, device, electronic device, and storage medium, which can improve the feedback training effect and stability in different subjects and different application scenarios.

[0004] Firstly, this application provides a feedback training method, which includes: The subjects' electroencephalogram (EEG) and physiological signals were acquired. The physiological signals were used to reflect the body's physiological state. The EEG and physiological signals had the same timestamp. The interaction index between EEG signals and physiological signals is calculated. The interaction index is used to characterize the degree of coupling between the subject's EEG signals and physiological signals. The higher the value of the interaction index, the higher the degree of coupling between the EEG signals and physiological signals. Interaction indicators are mapped to feedback parameters, and these parameters are presented to the subjects to guide them in self-regulating their physiological state based on these parameters.

[0005] As can be seen, this application ensures a precise temporal correspondence between central nervous system activity and peripheral autonomic physiological activity by simultaneously acquiring EEG and physiological signals and strictly aligning their timestamps, thus providing a data foundation for cross-modal analysis. Based on this, interaction indicators are calculated, transforming previously fragmented single-modal physiological data into core parameters that quantify the degree of brain-visceral coupling, overcoming the limitation of traditional single-modal feedback in comprehensively capturing the overall neuro-physiological state related to emotion. Subsequently, the interaction indicators are mapped and presented as feedback parameters in real time, constructing a multimodal closed-loop feedback mechanism that avoids training fluctuations caused by single signals being susceptible to scene interference. By using the interaction indicators between EEG and physiological signals as the core regulatory target, a global representation of complex neuro-physiological states is achieved, thereby improving training effectiveness and system stability in different subjects and application scenarios.

[0006] In a feasible example, physiological signals include electrocardiogram (ECG), respiratory, and gastric electrical signals; interaction metrics include a first interaction metric, a second interaction metric, and a third interaction metric; and the interaction metrics between EEG and physiological signals are calculated, including: Calculate the first interaction index between EEG and ECG signals; Calculate the second interaction index between electroencephalogram (EEG) signals and respiratory signals; Calculate the third interaction index between EEG signals and gastric electrical signals.

[0007] In this application, peripheral physiological signals are specifically decomposed into three typical visceral signals: electrocardiogram (ECG), respiratory signals, and gastric signals. These signals are mapped to the autonomic nervous activity characteristics of the cardiovascular, respiratory, and digestive systems, respectively. Based on this, the coupling calculation of EEG signals with these three types of visceral signals is performed independently to generate first, second, and third interaction indicators, forming a multi-dimensional brain-visceral synergistic regulation map. This parallel computing architecture overcomes the limitations of traditional single peripheral indicators, which are easily affected by individual physiological baseline differences or environmental noise interference. It can accurately quantify the independent linkage strength of the central nervous system to different visceral organs. Since emotions and overall physiological states are essentially collaborative responses of multi-organ autonomic neural networks, the parallel computation of multiple indicators effectively improves the robustness and comprehensiveness of feature representation. Using this multi-dimensional coupling state as the core regulatory target for feedback training, the system can adaptively match the neurophysiological response patterns of different subjects, thereby improving the stability of feedback training.

[0008] In a feasible example, the first interaction metric includes first interaction metric A, first interaction metric B, first interaction metric C, first interaction metric D, and first interaction metric E, wherein, The first interaction index A is used to characterize the response intensity of EEG signals to periodic pulses of ECG signals; The first interaction index B is used to characterize the degree of linear coordination between EEG signals and ECG signals; The first interaction index C is used to characterize the degree of phase locking between EEG signals and ECG signals; The first interaction index D is used to characterize the information flow and causal dominance direction between the characteristic sequences of EEG and ECG signals. The first interaction index E is used to characterize the degree to which the electrocardiogram (ECG) signal modulates the amplitude changes of the electroencephalogram (EEG) signal.

[0009] In this application, by subdividing the first interaction index into first interaction index A to first interaction index E, a quantitative representation of the brain-heart interaction index can be achieved by subdividing it into five orthogonal dimensions. The system comprehensively deconstructs the coupling mechanism between the central nervous system and the cardiovascular system from five levels: time-domain event response, frequency-domain synchronization characteristics, statistical correlation, causal dynamics, and bidirectional feedback mechanism. This multidimensional index system not only captures the neuroregulatory effect from the brain to the heart but also quantifies the physiological feedback effect from the heart to the brain, effectively overcoming the shortcomings of single coupling indexes that are easily interfered with by physiological noise or have one-sided representation. Using this refined and multidimensional brain-visceral coupling state as the core regulatory target of feedback training, the system can accurately map the overall neurophysiological dynamic characteristics related to the subject's emotions and adaptively match the physiological response patterns of different individuals, thereby improving the robustness and stability of feedback training in complex and ever-changing practical application scenarios.

[0010] In a feasible example, the second interaction metric includes second interaction metric A and second interaction metric B, wherein, The second interaction index A is used to characterize the degree of phase synchronization between EEG signals and respiratory signals; The second interaction index B is used to characterize the degree to which respiratory signals modulate changes in the amplitude of electroencephalogram (EEG) signals.

[0011] In this application, the second interaction index is refined into two orthogonal features: phase synchronization degree and amplitude modulation degree. The phase synchronization degree is used to accurately capture the dynamic alignment characteristics of the central nervous system rhythm and respiratory cycle in the time dimension to reflect the rhythmic regulation of the autonomic nervous system. The amplitude modulation degree is used to quantify the nonlinear effect of respiratory activity on EEG power to reveal the feedback effect of peripheral physiological state on central cortical excitability. This effectively overcomes the limitation of single coupling index being susceptible to individual differences in respiratory frequency or interference from environmental noise. The system can adaptively match the neuro-respiratory response patterns of different subjects, comprehensively characterize the overall physiological dynamics in the process of emotion regulation, and thus improve the robustness and stability of feedback training in complex and ever-changing practical application scenarios.

[0012] In a feasible example, the third interaction metric includes third interaction metric A and third interaction metric B, wherein, The third interaction index A is used to characterize the degree of signal coordination between electroencephalogram (EEG) signals and gastric electrical signals. The third interaction index, B, is used to characterize the degree to which gastric electrical signals modulate the amplitude changes of electroencephalogram (EEG) signals.

[0013] In this application, the third interaction index is refined into two orthogonal features: signal coordination degree and amplitude modulation degree. The signal coordination degree accurately quantifies the dynamic consistency between central nervous system rhythms and gastrointestinal electrical activity in the time-frequency domain, reflecting the overall regulatory level of the autonomic nervous system on digestive function. The amplitude modulation degree characterizes the nonlinear feedback effect of gastric electrical activity on brain electrical power, revealing the bottom-up physiological information transmission mechanism of the gut-brain axis. The combination of the two effectively overcomes the limitations of single indicators that are easily affected by individual differences in gastrointestinal peristalsis rhythms or motion artifacts. Using this interaction index as the core regulatory target for feedback training, the system can adaptively match the neuro-gastrointestinal response patterns of different subjects, comprehensively characterizing the overall physiological dynamics in the emotion regulation process, thereby improving the robustness and stability of feedback training in complex and ever-changing practical application scenarios.

[0014] In a feasible example, interaction metrics are mapped to feedback parameters, including: Feedback parameters are obtained by mapping the numerical magnitude, degree of change, and degree of deviation of the interaction indicator. The degree of change is used to characterize the degree of change between the interaction indicator and the previously generated interaction indicator. The degree of deviation is used to characterize the degree of deviation of the value of the interaction indicator from the preset numerical range. Feedback parameters include feedback intensity, feedback frequency, feedback rhythm, feedback continuity time, or feedback level.

[0015] In this application, the system comprehensively quantifies the absolute value of interactive indicators, the degree of change with preceding indicators, and the degree of deviation from preset value ranges. It accurately captures the instantaneous characteristics, temporal evolution patterns, and distance to the control target of the subject's brain-visceral coupling state. Based on this, the multidimensional mapping results are adaptively converted into diverse feedback parameters such as feedback intensity, frequency, rhythm, continuous time, or level. This achieves parameterization, personalization, and dynamic matching of feedback stimuli, effectively overcoming the shortcomings of traditional fixed threshold or single-dimensional feedback that are susceptible to physiological drift and noise interference. It enables subjects to make precise self-regulation based on real-time, dynamic, and clearly goal-oriented feedback parameters, which can improve the robustness and stability of cross-individual and cross-scenario feedback training.

[0016] In a feasible example, feedback parameters are presented to the subjects, including: Feedback parameters are presented to subjects based on target forms, which include one or more of the visual, auditory, and tactile forms.

[0017] In this application, by introducing a multimodal perception channel as the presentation carrier of feedback parameters and constructing a dynamic matching mechanism based on target form, the flexible switching or fusion of visual, auditory and tactile channels effectively overcomes the defects of fatigue, attention distraction and environmental noise interference that are easily caused by single sensory feedback, and can improve the perceptual salience and cognitive accessibility of feedback information.

[0018] Secondly, this application provides a feedback training device, which includes: The communication unit is used to acquire the subject's electroencephalogram (EEG) signals and physiological signals. The physiological signals are used to reflect the body's physiological state, and the EEG signals and physiological signals have the same timestamp. The processing unit is used to calculate the interaction index between EEG signals and physiological signals. The interaction index is used to characterize the degree of coupling between the subject's EEG signals and physiological signals. The higher the value of the interaction index, the higher the degree of coupling between the EEG signals and physiological signals. The processing unit is also used to map interaction metrics to feedback parameters; The communication unit is also used to present feedback parameters to the subject in order to guide the subject to self-regulate his / her physiological state based on the feedback parameters.

[0019] Thirdly, this application provides an electronic device including a processor, a memory, and a communication interface. The processor, memory, and communication interface are interconnected and perform communication with each other. The memory stores executable program code, the communication interface is used for wireless communication, and the processor is used to retrieve the executable program code stored in the memory and execute some or all of the steps described in any of the methods in the first aspect.

[0020] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements some or all of the steps described in the first aspect of this application.

[0021] Fifthly, this application provides a computer program product, including a computer program that, when processed and executed, implements some or all of the steps described in the first aspect of this application. The computer program product may be a software installation package. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a schematic diagram of the structure of a feedback training system provided in an embodiment of this application; Figure 2 This is a schematic diagram of the structure of a data acquisition module provided in an embodiment of this application; Figure 3 A flowchart illustrating a feedback training method provided in an embodiment of this application; Figure 4 A flowchart illustrating another feedback training method provided in an embodiment of this application; Figure 5 A flowchart illustrating yet another feedback training method provided in an embodiment of this application; Figure 6 A functional unit block diagram of a feedback training device provided in an embodiment of this application; Figure 7 A functional unit block diagram of another feedback training device provided in the embodiments of this application; Figure 8 This is a structural block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0024] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0025] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps is not limited to the steps listed, but may optionally include steps not listed, or may optionally include other steps inherent to these processes, methods, products, or apparatuses.

[0026] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0027] Currently, neurofeedback and biofeedback technologies are important techniques in the field of non-invasive functional regulation and rehabilitation training. Neurofeedback is typically based on brain activity signals such as electroencephalography (EEG), functional near-infrared spectroscopy (FIR), or functional magnetic resonance imaging (fMRI). It collects the subject's brain function status in real time and converts it into perceptible visual, auditory, or other forms of feedback, prompting the subject to learn to autonomously regulate specific brain activity patterns. Biofeedback, on the other hand, is usually based on peripheral physiological signals such as heart rate, heart rate variability, respiration, and electromyography (EMG). It is used to reflect autonomic nervous activity and the body's physiological state, and to assist in relaxation training, emotion regulation, and functional rehabilitation. Although existing feedback training techniques have been applied in cognitive training, emotion intervention, anxiety relief, sleep improvement, and rehabilitation therapy, most systems still belong to a single-modal feedback paradigm, usually revolving around central nervous activity or peripheral autonomic physiological activity, each forming a relatively independent feedback system.

[0028] In recent years, with the deepening of research in neuroscience and physiology, increasing evidence suggests a significant bidirectional coupling and dynamic interaction between brain activity and the rhythmic activity of visceral organs. For example, the brain is closely linked to the heart, respiratory system, and gastrointestinal system at the levels of rhythm, phase, synchronicity, and regulatory mechanisms. This brain-visceral interaction process is closely related to emotion regulation, stress response, autonomic nervous system balance, cognitive processing, and the occurrence and development of diseases. However, existing feedback training techniques mostly construct feedback mechanisms based on single-modal signals, failing to take the rhythmic coupling and dynamic interaction between the brain and visceral organs as the core regulatory object, thus making it difficult to comprehensively and accurately reflect the overall neurophysiological state of the body. In practice, existing feedback training systems often exhibit significant differences in training effects, insufficient stability, and limited individual adaptability in different subjects and application scenarios.

[0029] Based on this, this application uses brain-visceral coupling state as the core regulatory target for feedback training, which can more comprehensively characterize the overall neurophysiological state related to emotion, thereby improving the training effect in different subjects and different application scenarios.

[0030] Please see Figure 1 , Figure 1 This is a schematic diagram of the structure of a feedback training system provided in an embodiment of this application, as shown below. Figure 1 As shown, the feedback training system 100 includes a data acquisition module 101, a data processing module 102, and a feedback presentation module 103.

[0031] The data acquisition module 101 is used to synchronously acquire brain signals and visceral rhythm-related physiological signals of the subject during training. Brain signals are electroencephalogram (EEG) signals, and visceral rhythm-related physiological signals include one or more of the following: electrocardiogram (ECG), respiratory signals, and electrogastric signal (EGG). This module can acquire corresponding signals through sensors placed on the subject's head, chest, abdomen, and / or abdominal surface, achieving time-synchronous acquisition of multimodal signals.

[0032] Please see Figure 2 , Figure 2 This is a schematic diagram of the structure of a data acquisition module provided in an embodiment of this application, as shown below. Figure 2 As shown, the data acquisition module 101 acquires EEG signals and visceral rhythm-related physiological signals by placing EEG acquisition electrodes on the subject's head and deploying ECG, respiration, and gastric electroencephalogram sensors on the chest and / or abdomen. After the subject is fitted with the electrodes, the signal quality of each channel is checked to confirm that the electrode contact is stable, the sensors are working properly, and each acquisition channel can continuously output valid signals. Subsequently, the data acquisition module 101 transmits the synchronously acquired EEG signals and physiological signals to the data processing module 102 via wired or wireless means.

[0033] The data processing module 102 is used to preprocess the received EEG signals and physiological signals, extract features, calculate interactive indicators, and map feedback parameters. It then transmits the feedback parameters to the feedback presentation module 103 via wired or wireless means.

[0034] The feedback presentation module 103 is used to output feedback parameters to the subject in a visual, auditory and / or tactile manner.

[0035] Visual feedback can be changes in image brightness, color, or graphic size. For example, real-time calculated interactive metrics can be mapped to bar charts, graphs, dashboards, progress bars, or score changes on the screen; or they can be mapped to dynamic changes such as zooming in, zooming out, moving, rotating, brightening, darkening, sharpening, or blurring of a graphic target; or they can be designed as feedback from natural scenes, such as changes in sky brightness, ocean waves, plant growth, light spot aggregation, or changes in the movement state of a virtual character, allowing subjects to complete self-regulation training by observing real-time changes in visual objects.

[0036] Auditory feedback can manifest as changes in pitch, volume, background music, or voice prompts. For example, when the target interaction indicator moves toward the target range, the pitch can be increased, harmony enhanced, noise reduced, or a positive prompt outputted; when the target interaction indicator deviates from the target range, the pitch can be lowered, the music's fluency reduced, or a corrective voice prompt triggered. Audio feedback is particularly common in respiratory biofeedback and relaxation training because it does not require the subject to continuously look at a screen, making it suitable for use in conjunction with guided breathing rhythms.

[0037] Tactile feedback can manifest as changes in vibration intensity, vibration frequency, or pulse rhythm. It is commonly used in scenarios where continuous viewing of a visual interface is inconvenient, and is often combined with visual feedback to form a visual-tactile bimodal feedback system. Existing research has compared the roles of visual and vibrational feedback in brain-computer interfaces and respiratory rhythm training, indicating that tactile feedback can serve as an effective alternative or supplementary method.

[0038] Through the above feedback presentation method, subjects can perceive the current brain-visceral rhythm interaction state in real time and make autonomous adjustments accordingly.

[0039] In this application, the data processing module 102 acquires the electroencephalogram (EEG) and physiological signals of the subject collected by the data acquisition module 101. The physiological signals reflect the body's physiological state, and the EEG and physiological signals have the same timestamp. It calculates an interaction index between the EEG and physiological signals, which characterizes the degree of coupling between them; a higher value indicates a higher degree of coupling. The interaction index is mapped to feedback parameters, which are then presented to the subject through the feedback presentation module 103 to guide the subject in self-regulating their physiological state based on these parameters. This feedback training, using the coupling state between EEG and physiological signals as the core regulatory target, can more comprehensively characterize the overall neurophysiological state related to emotion, thereby improving the feedback training effect in different subjects and application scenarios.

[0040] Based on this, the embodiments of this application provide a feedback training method, and the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0041] Example 1: The main process of the feedback training method will be described below.

[0042] Please see Figure 3 , Figure 3 This is a flowchart illustrating a feedback training method provided in an embodiment of this application. The method is applied to the aforementioned data processing module, such as... Figure 3 As shown, the method includes the following steps.

[0043] Step S301: Acquire the subject's electroencephalogram (EEG) signals and physiological signals.

[0044] Physiological signals are used to reflect the body's physiological state, and the timestamps of EEG signals and physiological signals are the same.

[0045] Electroencephalogram (EEG) signals can be continuous waveform data reflecting the electrical activity of neuronal populations in the central nervous system, and can be used to provide neural activity representations of emotional cognitive processing and attention regulation. For example, EEG signals can include, but are not limited to, signal components classified by frequency band, such as delta waves, theta waves, alpha waves, beta waves, and gamma waves. Physiological signals can be continuous physiological parameter data reflecting the activity state of the peripheral autonomic nervous system, and can be used to provide peripheral representations of emotional arousal and changes in autonomic nervous tension. Timestamps can be precise time identifiers marking the time of data acquisition, and can be used to ensure a strict correspondence between multimodal data on the timeline. Subjects can be individuals participating in feedback training and receiving systematic intervention, and can serve as the implementers of feedback regulation and the subjects of effect verification.

[0046] After acquiring EEG and physiological signals, data preprocessing can be performed. For example, preprocessing of EEG signals includes: bandpass filtering of the acquired raw EEG signals to preserve the target EEG frequency band and suppress low-frequency drift and high-frequency noise; a bandpass filtering range of approximately 0.5Hz to 45Hz can be used. For low-frequency drift in the EEG signals caused by changes in electrode contact, slow body movements, etc., moving average removal is used to improve the stability of subsequent feature calculations. Online identification and suppression of eye movement artifacts, blinking artifacts, electromyography artifacts, and sudden movement artifacts can be achieved through amplitude threshold discrimination, signal slope discrimination, moving variance detection, reference channel correction, or adaptive filtering methods. For time windows where artifacts are obvious, removal or interpolation can be performed.

[0047] Preprocessing of electrocardiogram (ECG) signals within physiological signals includes: bandpass filtering of the raw ECG signal to retain effective components such as the QRS complex and suppress low-frequency drift and high-frequency noise; a bandpass filtering range of approximately 0.5 Hz to 40 Hz can be used. Baseline drift caused by respiration, body movement, and electrode contact changes is removed using high-pass filtering, median filtering, morphological filtering, or sliding baseline correction. Online R-peak detection is performed on the preprocessed ECG signal to extract adjacent beat interval sequences; thresholding, derivative methods, template matching, or adaptive peak detection methods can be used to identify the R-peak position. Abnormal, missed, and falsely detected heartbeats are identified and corrected online; anomalies can be determined based on the range of adjacent beat interval changes, statistical outlier thresholds, or local trends, and the beat interval sequences are corrected through interpolation or smoothing to ensure the continuity and stability of subsequent rhythm analysis.

[0048] Preprocessing of respiratory signals in physiological signals includes: bandpass filtering of the raw respiratory signal; retaining low-frequency respiratory components and suppressing high-frequency noise when baseline drift needs to be suppressed simultaneously; a bandpass filtering range of 0.05–2Hz or 0.05–3Hz can be used. For slow drift caused by sensor loosening, body movement, or posture changes, a sliding window method is used to remove trends or normalize the baseline of the respiratory signal, resulting in a baseline-removed respiratory signal. Peak and trough detection is performed on the preprocessed respiratory signal to determine parameters such as the onset of inspiration, the onset of expiration, the respiratory cycle, respiratory rate, and respiratory amplitude. Preferably, the zero-crossing method, local extremum method, or threshold method can be used for period identification. Abnormal respiratory segments caused by coughing, talking, body movement, etc., can be identified through amplitude mutation detection, period anomaly detection, or local variance detection, and then eliminated or interpolated.

[0049] Preprocessing of gastric electroencephalogram (GEG) signals in physiological signals includes: Since GEG signals primarily consist of low-frequency slow wave components, low-frequency bandpass filtering is applied to the raw GEG signal to preserve slow wave activity and suppress high-frequency noise and non-gastric interference. The selectable bandpass range is 0.015–0.25 Hz. For common slow drift phenomena in GEG signals, trend removal, moving mean correction, or high-pass correction methods are used. Abnormal segments caused by large-amplitude body movements, abdominal muscle contractions, or sudden interference are identified through amplitude thresholds, local energy, slope abrupt changes, or spectral anomalies, and these segments are then removed, labeled, or interpolated. Sliding window segmentation analysis is performed on the preprocessed GEG signal to extract gastric slow wave rhythm-related parameters, such as dominant frequency, amplitude, rhythm stability, and effective slow wave bands, for subsequent brain-gastric coupling analysis.

[0050] Step S302: Calculate the interaction index between EEG signals and physiological signals.

[0051] Interaction metrics can be numerical parameters that quantify the strength of the dynamic correlation between EEG signals and physiological signals, transforming fragmented single-modal data into core parameters characterizing the degree of brain-visceral coupling. In this embodiment, interaction metrics can be based on aligned bimodal data sequences, extracting correlation features through frequency domain coherence analysis, mutual information calculation, or phase synchronization index algorithms. For example, interaction metrics can include, but are not limited to, one or more of the following, categorized by computational dimension: time-domain cross-correlation coefficient, frequency-domain coherence function value, and nonlinear mutual information.

[0052] The degree of coupling can be an abstract attribute describing the strength of the coordinated linkage between central nervous activity and peripheral autonomic physiological activity, and can reflect the level of integration of the overall neurophysiological state related to emotions.

[0053] Step S303: Map the interaction indicators to feedback parameters and present the feedback parameters to the subjects to guide them to self-regulate their physiological state based on the feedback parameters.

[0054] The feedback parameters can be real-time output information of interactive index values ​​converted into human-perceptible modalities, enabling the construction of a multimodal closed-loop feedback mechanism that allows subjects to self-regulate based on the integrated coupling state. In this embodiment, the feedback parameters can convert interactive index values ​​into physical quantities such as visual brightness, auditory frequency, or tactile vibration intensity according to a preset mapping rule.

[0055] Mapping interactive metrics to feedback parameters can be achieved by converting the numerical values ​​of the interactive metrics into control commands for specific physical quantities, based on a preset monotonic mapping function or piecewise linear rules. Similarly, artificial intelligence models can be introduced to learn the mapping relationship between the magnitude, degree of change, and deviation of the interactive metrics and the feedback parameters. For example, mapping interactive metrics to feedback parameters can be achieved by using proportional-integral mapping to generate visually gradual animations or by triggering audio cues of different frequencies through threshold-based piecewise mapping. This transforms abstract mathematical parameters into intuitive and perceptual stimuli, reducing the cognitive load on the subject and improving regulatory directionality.

[0056] Feedback parameters are presented to subjects via a display terminal, speaker, or haptic actuator, outputting mapped physical stimuli while maintaining a refresh rate consistent with the update frequency of the interactive indicators. Furthermore, guiding subjects to self-regulate their physiological state based on the feedback parameters can be achieved through techniques such as diaphragmatic breathing combined with visual signal synchronization training, and mindfulness meditation combined with auditory signal frequency tracking. This can stimulate the subjects' endogenous regulatory abilities, causing the values ​​of the interactive indicators to converge towards the target range, thus completing the training intervention.

[0057] In this embodiment, by synchronously acquiring EEG signals and physiological signals and strictly aligning their timestamps, the precise temporal correspondence between central nervous system activity and peripheral autonomic physiological activity is ensured, providing a data foundation for cross-modal analysis. Based on this, interaction indicators are calculated, transforming previously fragmented single-modal physiological data into core parameters that quantify the degree of brain-visceral coupling, overcoming the limitation of traditional single-modal feedback in comprehensively capturing the overall neuro-physiological state related to emotion. Subsequently, the interaction indicators are mapped and presented as feedback parameters in real time, constructing a multimodal closed-loop feedback mechanism that avoids training fluctuations caused by single signals being susceptible to scene interference. By using the interaction indicators between EEG signals and physiological signals as the core regulatory target, a global representation of complex neuro-physiological states is achieved, thereby improving training effectiveness and system stability in different subjects and application scenarios.

[0058] In Example 2, the physiological signals include electrocardiogram (ECG), respiratory signals, and gastric electrical signals. The interaction indicators include a first interaction indicator, a second interaction indicator, and a third interaction indicator. The feedback training method will be described in detail below based on the calculation details of the interaction indicators.

[0059] Please see Figure 4 , Figure 4 This is a flowchart illustrating another feedback training method provided in an embodiment of this application. This method is applied to the aforementioned data processing module, such as... Figure 4 As shown, the method includes the following steps.

[0060] Step S401: Acquire the subject's electroencephalogram (EEG) signals and physiological signals.

[0061] Step S402: Calculate the first interaction index between the EEG signal and the ECG signal.

[0062] The electrocardiogram (ECG) signal can be continuous waveform data reflecting cardiac electrophysiological activity and the autonomic nervous system regulation state of the cardiovascular system, and can be used to provide a peripheral characterization of the balance between sympathetic and parasympathetic nervous system tension in the cardiovascular system. In this embodiment, the ECG signal can be acquired through surface electrodes to collect the potential difference changes generated during the depolarization and repolarization of myocardial cells.

[0063] The first interaction index can be a specific numerical parameter that quantifies the strength of the dynamic correlation between EEG and ECG signals, and can be used to accurately map the intensity of the central nervous system's regulation of the cardiovascular system. For example, the first interaction index between EEG and ECG signals can be calculated by using a sliding window to calculate the coherence between the EEG frequency band and the heart rate variability frequency band, and by using a mutual information algorithm to assess the nonlinear dependence between the EEG waveform and the RR interval sequence. This allows for the independent quantification of the intensity of the central nervous system's regulation of the cardiovascular system, avoiding feature dilution caused by the mixing of signals from multiple organs.

[0064] In one embodiment, the first interaction index includes first interaction index A, first interaction index B, first interaction index C, first interaction index D, and first interaction index E, wherein first interaction index A is used to characterize the response intensity of the EEG signal to the periodic pulses of the ECG signal; first interaction index B is used to characterize the degree of linear coordination between the EEG signal and the ECG signal; first interaction index C is used to characterize the degree of phase locking between the EEG signal and the ECG signal; first interaction index D is used to characterize the information flow direction and causal dominance direction between the feature sequences of the EEG signal and the ECG signal; and first interaction index E is used to characterize the degree of modulation of the amplitude change of the EEG signal by the ECG signal.

[0065] The first interactive index A can be a numerical parameter that quantifies the response of brain electrical activity to periodic events triggered by electrocardiograms (ECGs). It can be used to capture the transient driving effect of cardiovascular rhythms on the central nervous system, such as the heart evoked potential (HEP) index. In this embodiment, the first interactive index A can locate the time points of periodic events in the ECG signal, extract the EEG segments within the corresponding time window, and calculate the relative rate of change of energy or amplitude. Furthermore, characterizing the response intensity of the EEG signal to the periodic pulses of the ECG signal can be achieved by using event-related potential analysis to extract the EEG amplitude increment after the R wave and using time-frequency analysis to calculate the spectral energy concentration after the periodic event is triggered, thereby quantifying the transient driving effect of peripheral cardiovascular rhythms on the central nervous system.

[0066] Furthermore, the first interaction index A between the electroencephalogram (EEG) signal and the electrocardiogram (ECG) signal is calculated, including: Obtain the time point sequence corresponding to all R-wave peaks in the electrocardiogram (ECG) signal; using each R-wave peak time point as the time zero point, extract EEG signal segments of a first preset time length to obtain multiple EEG signal segments; extract the EEG signal of the first preset time interval before the time zero point of each EEG signal segment as the baseline signal, calculate the average voltage value of the baseline signal, and subtract the average voltage value from all voltage values ​​within the EEG signal segment to obtain the baseline-corrected EEG signal segment; superimpose all baseline-corrected EEG signal segments at the same time point and calculate the arithmetic mean to obtain the cardiac evoked potential waveform; extract the maximum voltage amplitude within the second preset time interval after the time zero point from the cardiac evoked potential waveform as the first interactive index A.

[0067] First, the peak value of the R-wave with the largest amplitude and most prominent characteristics in the ECG signal is located, and its corresponding time point is used as the precise time trigger reference (time zero point), because the R-wave represents the moment of strongest electrical activity of the periodic pulse of the ECG signal. Using each R-wave time point as zero point, a segment of EEG signal from 200ms to 600ms is extracted. This time window is set because there is a physical delay in the response of EEG signal to ECG signal events, and this interval can completely cover the typical transient response latency. The EEG signal from 100ms to 200ms before the zero point is extracted as the baseline, its average voltage is calculated, and subtracted from subsequent segments. This mathematically eliminates the DC offset and low-frequency drift of the EEG signal itself, ensuring that the extracted voltage change is purely a relative amplitude induced by the ECG signal event. All baseline-corrected EEG signals were segmented, strictly aligned on the time axis, and then arithmetically averaged. This is because background EEG noise is randomly distributed, while the phase-locked response to the ECG signal is fixed. The superposition and averaging effectively cancels out random noise using the law of large numbers, highlighting the true evoked waveform. Finally, the maximum voltage amplitude within a 250ms to 350ms window was extracted from the averaged waveform. This allows for direct quantification of the average response intensity of the EEG signal to a single ECG signal pulse using the maximum amplitude.

[0068] It can be seen that by employing signal processing techniques such as time locking, baseline correction, and superposition averaging, the transient components of the EEG signal that are strictly phase-locked with the R-wave of the ECG signal are accurately extracted in the time domain, effectively suppressing asynchronous random background EEG noise and baseline drift. This provides a high signal-to-noise ratio time-domain characteristic benchmark for the feedback training system, enabling the system to objectively quantify the response intensity of the EEG signal to the periodic pulses of the ECG signal, thereby accurately assessing the changes in the emotional arousal of the subject during training.

[0069] The first interaction index B can be a numerical parameter characterizing the strength of the statistically significant linear correlation between EEG and ECG signals, reflecting the synchronous fluctuation characteristics of the central nervous system and cardiovascular system in macroscopic trends. In an exemplary embodiment, the first interaction index B can be used to standardize the two signal sequences and calculate the linear correlation coefficient or regression fit goodness of fit of numerical changes. For example, characterizing the degree of linear coordination between EEG and ECG signals can be achieved by using the Pearson correlation coefficient to assess the consistency of the trends of the two signals, using the least squares method to fit a linear regression model and extract the coefficient of determination, thereby reflecting the statistical synchronicity of the central nervous system and cardiovascular system in macroscopic fluctuation trends.

[0070] Furthermore, the first interaction index B between the EEG and ECG signals is calculated, including: The EEG signal is segmented according to a sliding time window with a second preset time length and a first sliding step size. A Fast Fourier Transform is performed on the EEG signal within each segment to calculate the power spectral density of the first preset frequency band. The segments are then arranged in chronological order to obtain the EEG power time series. Based on the R-wave peak time point sequence of the ECG signal, the time difference between adjacent R-wave peaks is calculated to obtain the RR interval sequence. The RR interval sequence is resampled along the same time axis as the sliding time window, and the root mean square error of the RR interval within each first preset time window is calculated. These sequences are then arranged in chronological order to obtain the heart rate variability time series. The Pearson correlation coefficient between the EEG power time series and the heart rate variability time series is calculated as the first interaction index B.

[0071] The EEG signal was segmented using a sliding time window with a length of 5 seconds and a step size of 1 second. A Fast Fourier Transform (FFT) was performed on each segment to calculate the power spectral density in the 8Hz to 13Hz frequency band. The sliding window ensured temporal resolution, and the FFT converted the time-domain signal into frequency-domain energy, forming a power time series reflecting the dynamic changes in energy within a specific frequency band of the EEG signal. The time difference between adjacent R-wave peaks in the ECG signal was calculated to obtain the RR interval sequence, which records the variation in the duration of adjacent beat cycles in the ECG signal. The discrete RR interval sequences were resampled and aligned along the same time axis as the EEG sliding window, and the root mean square error (RMSSD) of the RR intervals within each 5-second window was calculated. The RMSSD mathematically quantifies the intensity of fluctuations in the ECG signal cycle within a local time window, forming a heart rate variability time series. Finally, the Pearson correlation coefficient between the EEG power time series and the heart rate variability time series was calculated. This coefficient quantifies the degree of linear correlation between two time series in the macroscopic time domain by calculating the ratio of their covariance to standard deviation.

[0072] This transforms the discrete RR intervals of the ECG signal into a continuous sequence aligned with the frequency domain power time axis of the EEG signal, enabling linear covariance calculation of the two heterogeneous signals at a macroscopic time domain scale. This allows the system to monitor the synchronous evolution trend of EEG signal frequency band energy and ECG signal rhythm fluctuations in real time, providing a macroscopic quantitative basis for assessing the overall fatigue and relaxation state of subjects and guiding the dynamic adjustment of training programs.

[0073] The first interaction index C can be a numerical parameter that quantifies the synchronization stability between the EEG oscillation phase and the ECG rhythm phase, and can be used to reveal the rhythmic coordination mechanism of the central nervous system and the cardiovascular system at the frequency domain level. In this embodiment, the first interaction index C can extract the instantaneous phase sequence of the two signals and calculate the concentration of the phase difference distribution or the phase lock value. For example, characterizing the phase lock degree of the EEG signal and the ECG signal can be achieved by using Hilbert transform to extract the instantaneous phase and calculate the phase lock value, and using circular statistics to evaluate the skewness and kurtosis of the phase difference distribution, thereby revealing the rhythmic synchronization stability of the central nervous system and the cardiovascular system within a specific frequency band.

[0074] Furthermore, the first interaction index C between the EEG signal and the ECG signal is calculated, including: The EEG signal is bandpass filtered in a second preset frequency band, and then Hilbert transform is performed on the filtered EEG signal to extract the instantaneous phase time series of the EEG signal. The RR interval sequence is interpolated multiple times using the same time sampling rate as the EEG signal to obtain a continuous RR interval time series. The RR interval time series is then subjected to Hilbert transform to extract the instantaneous phase time series of the heart rhythm. The phase difference between the instantaneous phase time series of the EEG signal and the instantaneous phase time series of the heart rhythm at each time point is calculated, and the arithmetic mean of the cosine values ​​of the phase differences at all time points is taken as the first interaction index C.

[0075] The EEG signal was bandpass filtered from 8Hz to 13Hz to remove interference outside the target frequency band. A Hilbert transform was then performed to construct an analytic signal. By calculating the arctangent values ​​of the imaginary and real parts of the analytic signal, the instantaneous phase time series of the EEG signal at each sampling point was extracted. Since the RR interval sequence is a discrete event, it was first resampled using cubic spline interpolation to a continuous time series with the same sampling rate as the EEG signal, ensuring time axis alignment. Then, a Hilbert transform was performed on this continuous sequence to extract the instantaneous phase time series of the ECG rhythm. The phase difference between the two instantaneous phase sequences at each time point was calculated, and the cosine value of all phase differences was obtained and the arithmetic mean was calculated. A cosine value of 1 indicates complete in-phase, -1 indicates complete out-of-phase, and 0 indicates random phase. The arithmetic mean (i.e., the phase-locked value) mathematically quantifies the degree of convergence and locking of the two signal phases.

[0076] This method resamples the non-stationary RR interval of the ECG signal into a continuous signal with the same sampling rate as the EEG signal, achieving instantaneous phase alignment and lock-in calculation between the two on the microscopic time axis. This enables the system to sensitively capture the phase-carrying effect between the EEG and ECG signals, providing a highly sensitive core assessment parameter for measuring the synergy of brain-heart rhythms in subjects during deep relaxation training.

[0077] The first interaction index D can be a numerical parameter characterizing the causal driving direction and intensity between the characteristic sequences of EEG and ECG signals. It can be used to clarify the unidirectional dominant path of central nervous system regulation of the cardiovascular system or cardiovascular feedback center. In a specific embodiment, the first interaction index D can be based on an autoregressive model or information theory method to evaluate the difference in predictive ability between the two signal characteristic sequences to determine the dominant path. For example, characterizing the information flow direction and causal dominant direction between the characteristic sequences of EEG and ECG signals can be achieved by using Granger causality tests to calculate the bidirectional prediction gain ratio and using the transfer entropy algorithm to evaluate the net flow of nonlinear information transmission, thereby clarifying the unidirectional driving path and dominant control source in the brain-heart interaction process.

[0078] Furthermore, the first interaction index D between the EEG signal and the ECG signal is calculated, including: A bivariate autoregressive model was constructed with EEG power time series as the first variable and heart rate variability time series as the second variable. The coefficient matrix of the autoregressive model was solved using the least squares method. Based on the coefficient matrix and the residual covariance matrix of the autoregressive model, the transfer entropy values ​​from the first variable to the second variable and from the second variable to the first variable were calculated, and the two transfer entropy values ​​were used as the first interaction index D.

[0079] First, a bivariate autoregressive (VAR) model is constructed using the EEG signal power time series as the first variable and the ECG signal heart rate variability time series as the second variable. This model assumes that the signal value at the current moment can be predicted by a linear combination of the two variable values ​​from several past moments. Second, the lag order is set to 2 to 5 (covering the typical time delay of signal interaction), and the coefficient matrix of the model is solved using the least squares method to minimize the sum of squared residuals between the model's predicted values ​​and the actual signal values. Finally, the bidirectional transfer entropy value is calculated based on the solved coefficient matrix and residual covariance matrix. Transfer entropy, based on the concept of conditional entropy in information theory, mathematically quantifies the amount of nonlinear information transfer from the first variable to the second variable and from the second variable to the first variable by comparing "the reduction in uncertainty of the future state of variable B after knowing the past state of variable A".

[0080] By constructing a time-series prediction model and quantifying the asymmetry of information transmission, the nonlinear information flow between the characteristic sequences of EEG and ECG signals is mathematically analyzed. This provides a clear directional intervention basis for neurofeedback systems, enabling the system to automatically determine whether the current state is dominated by EEG signals or ECG feedback, and thus intelligently switch training strategies to prioritize regulating EEG or guiding ECG rhythms.

[0081] The first interaction index E can be a numerical parameter that quantifies the modulation effect of cardiac activity on the amplitude envelope of the electroencephalogram (EEG) signal, and can be used to characterize the periodic shaping effect of peripheral cardiovascular rhythms on central nervous system excitability. In this embodiment, the first interaction index E can analyze the periodic fluctuation characteristics of the EEG amplitude envelope within the cardiac rhythm cycle and calculate the modulation depth or modulation index. For example, characterizing the degree of modulation of the EEG signal amplitude variation by the cardiac signal can be achieved by using a modulation index algorithm to assess the fluctuation depth of the EEG amplitude with the heartbeat cycle, and by using a synchronization likelihood method to analyze the periodic influence of the cardiac phase on the EEG power, thereby characterizing the periodic shaping effect of peripheral cardiovascular rhythms on central nervous system excitability.

[0082] Furthermore, the first interaction index D between the EEG signal and the ECG signal is calculated, including: Hilbert transform is applied to the RR interval time series of the electrocardiogram (ECG) signal to extract the instantaneous phase time series of the RR interval as a low-frequency phase variable. Bandpass filtering of the EEG signal in a third preset frequency band is performed, followed by Hilbert transform to extract the instantaneous amplitude envelope time series of the EEG signal as a high-frequency amplitude variable. The value range of the low-frequency phase variable is divided into multiple equal phase intervals, and the average amplitude of the high-frequency amplitude variable in each phase interval is calculated. The multiple average amplitudes are normalized to obtain a probability distribution, and the information entropy of the probability distribution is calculated. The information entropy is then normalized by dividing the information entropy by the natural logarithm of the maximum possible information entropy, and the resulting value is used as the first interaction index D.

[0083] First, a Hilbert transform was performed on the continuous ECG signal RR interval time series to extract its instantaneous phase time series, which served as the phase variable for modulating the low-frequency rhythm. Second, the EEG signal was bandpass filtered from 8Hz to 13Hz and then subjected to a Hilbert transform to extract the modulus of its analytic signal (i.e., the instantaneous amplitude envelope), which served as the modulated high-frequency amplitude variable. The range of the low-frequency phase variable (0 to 2π) was divided into 18 equal phase intervals, and the average amplitude of the high-frequency amplitude variable falling into each phase interval was statistically analyzed. This step established a statistical mapping relationship between the low-frequency phase period and the high-frequency amplitude fluctuation. Finally, the 18 average amplitudes were normalized to obtain a probability distribution, and the information entropy of this distribution was calculated. If the amplitude is uniformly distributed across all phase intervals, the information entropy is maximized (no coupling); if the amplitude is concentrated in a specific phase, the information entropy decreases (strong coupling). Normalizing the information entropy by dividing it by the natural logarithm of the maximum possible information entropy yields a value that rigorously quantifies the modulation depth of the low-frequency phase on the high-frequency amplitude.

[0084] By employing cross-band statistical mapping, a nonlinear probability distribution model is established between the low-frequency phase cycle of the electrocardiogram (ECG) signal and the high-frequency amplitude fluctuation of the electroencephalogram (EEG) signal, quantifying the modulation depth of the low-frequency rhythm on the high-frequency oscillation. This enables the system to assess the regularity of the EEG signal amplitude fluctuations with the ECG signal cycle. The attenuation or enhancement of this coupling index can be directly used as a criterion for the disorder or recovery of brain-heart signal interaction, and can be used for the precise tracking of the therapeutic effects of long-term training.

[0085] It can be seen that by subdividing the first interaction index into first interaction index A to first interaction index E, a quantitative representation of the brain-heart interaction index can be achieved by subdividing it into five orthogonal dimensions. The system comprehensively deconstructs the coupling mechanism between the central nervous system and the cardiovascular system from five levels: time-domain event response, frequency-domain synchronization characteristics, statistical correlation, causal dynamics, and bidirectional feedback mechanism. This multidimensional index system not only captures the neuroregulatory effect from the brain to the heart but also quantifies the physiological feedback effect from the heart to the brain, effectively overcoming the shortcomings of single coupling indexes that are easily interfered with by physiological noise or have one-sided representation. Using this refined and multidimensional brain-visceral coupling state as the core regulatory target of feedback training, the system can accurately map the overall neurophysiological dynamic characteristics related to the subject's emotions and adaptively match the physiological response patterns of different individuals, thereby improving the robustness and stability of feedback training in complex and ever-changing practical application scenarios.

[0086] Step S403: Calculate the second interaction index between the electroencephalogram (EEG) signal and the respiratory signal.

[0087] The respiratory signal can be continuous physiological parameter data reflecting changes in thoracic cavity volume and the autonomic nervous drive pattern of the respiratory system, and can be used to provide a dynamic representation of the respiratory center rhythm and peripheral gas exchange status. In this embodiment, the respiratory signal can be recorded by a chest and abdominal belt sensor or an impedance respiratory monitoring device to record airflow or volume changes during the respiratory cycle.

[0088] The second interaction index can be a dedicated numerical parameter that quantifies the strength of the dynamic correlation between EEG and respiratory signals, and can be used to accurately map the intensity of the central nervous system's regulation of respiratory rhythm. In one specific embodiment, the second interaction index can be extracted based on aligned EEG and respiratory data sequences using a cross-modal coupling algorithm to extract brain-respiratory linkage features. For example, calculating the second interaction index between EEG and respiratory signals can be achieved by using a phase-locked value algorithm to analyze the phase synchronization between EEG oscillations and the respiratory cycle, and by using a dynamic time warping method to evaluate the morphological matching degree between the EEG envelope and the respiratory volume curve. This allows for the independent quantification of the intensity of the central nervous system's regulation of respiratory rhythm, improving the directivity of respiratory-related autonomic neural feedback.

[0089] In one embodiment, the second interaction index includes a second interaction index A and a second interaction index B, wherein the second interaction index A is used to characterize the degree of phase synchronization between the electroencephalogram (EEG) signal and the respiratory signal; and the second interaction index B is used to characterize the degree of modulation of the amplitude change of the EEG signal by the respiratory signal.

[0090] The second interaction index A can be a numerical parameter that quantifies the dynamic alignment stability between the EEG oscillation phase and the respiratory cycle phase, and can be used to capture the synergistic characteristics of central nervous system rhythms and respiratory activity in the time dimension. In this embodiment, the second interaction index A can extract the instantaneous phase sequence of the two signals and calculate the concentration of the phase difference distribution or the phase lock value.

[0091] Phase synchronization can be an abstract attribute describing the fixed or stable difference relationship between the phase trajectories of two periodic signals on the time axis. It can be used as an intrinsic representation dimension of index A and reflects the stability of rhythm coordination. Furthermore, the phase synchronization between EEG and respiratory signals can be characterized by extracting the instantaneous phase using Hilbert transform and calculating the phase lock value, and by using circular statistics to evaluate the skewness and kurtosis of the phase difference distribution. This can reveal the time alignment stability of the central nervous system and respiratory rhythms within a specific frequency band.

[0092] Furthermore, the second interaction index A between the electroencephalogram (EEG) signal and the respiratory signal is calculated, including: The EEG signal is bandpass filtered in the fourth preset frequency band and then subjected to Hilbert transform to extract the instantaneous phase time series of the EEG signal; the respiratory signal is subjected to Hilbert transform to extract the instantaneous phase time series of the respiratory signal; the phase difference between the instantaneous phase time series of the EEG signal and the instantaneous phase time series of the respiratory signal at each time point is calculated, and the arithmetic mean of the cosine values ​​of the phase differences at all time points is obtained as the second interaction index A.

[0093] First, the EEG signal was bandpass filtered from 8Hz to 13Hz, and its instantaneous phase time series was extracted using Hilbert transform to obtain the periodic phase changes of the EEG rhythm. Second, the respiratory signal was directly subjected to Hilbert transform to extract its instantaneous phase time series. Since the respiratory signal itself is a continuous periodic physical waveform, interpolation and resampling are unnecessary as with the RR interval in ECG, directly preserving its original physical fluctuation phase. Finally, the phase difference between the two instantaneous phase sequences at each time point was calculated, and the cosine value of all phase differences was obtained and the arithmetic mean was calculated. This calculation process mathematically evaluates the degree of clustering between the EEG signal phase and the respiratory signal phase on the time axis; the closer the value is to 1, the more synchronized their periodic fluctuations are.

[0094] This approach directly achieves instantaneous phase alignment between specific frequency bands of EEG signals and the original cycle of respiratory signals without the need for interpolation and resampling, preserving the most authentic physical fluctuation characteristics of respiratory signals. Since respiratory rhythm is significantly influenced by subjective consciousness, this indicator can intuitively quantify the phase-carrying effect of the subject's active respiratory regulation on EEG signals, making it the most sensitive real-time effect assessment parameter for respiratory biofeedback training.

[0095] The second interaction index B can be a numerical parameter that quantifies the periodic modulation effect of respiratory rhythm on the power or amplitude of EEG signals, and can be used to characterize the nonlinear feedback effect of peripheral respiratory activity on the excitability of the central cortex. In this embodiment, the second interaction index B can be used to extract the segmented envelope of the EEG signal based on the respiratory cycle and calculate the matching degree or modulation index between the envelope fluctuation and the respiratory rhythm.

[0096] Amplitude variation can describe the physical characteristics of signal peak and trough height or energy intensity fluctuations over time, serving as the basis for calculating index B and providing a quantitative basis for the fluctuations in central nervous system excitability. Modulation degree can describe the intensity attribute of the periodic shaping effect of one periodic signal on the amplitude or energy of another signal, serving as a core output feature of index B and characterizing the peripheral-to-internal reverse driving effect. Furthermore, characterizing the modulation degree of respiratory signals on EEG amplitude variations can be achieved by using modulation index algorithms to assess the depth of EEG amplitude fluctuations with the respiratory cycle and by using synchronization likelihood methods to analyze the periodic influence of respiratory phase on EEG power, thereby characterizing the periodic shaping effect of peripheral respiratory activity on central nervous system excitability.

[0097] Furthermore, the second interaction index B between the electroencephalogram (EEG) signal and the respiratory signal is calculated, including: The instantaneous phase time series of the respiratory signal is used as the low-frequency phase variable; the EEG signal is bandpass filtered in the fifth preset frequency band and subjected to Hilbert transform to extract the instantaneous amplitude envelope time series of the EEG signal as the high-frequency amplitude variable; the value range of the low-frequency phase variable is divided into multiple equal phase intervals, and the average amplitude of the high-frequency amplitude variable in each phase interval is calculated; the multiple average amplitudes are normalized to obtain a probability distribution, the information entropy of the probability distribution is calculated, and the information entropy is normalized by dividing the information entropy by the natural logarithm of the maximum possible information entropy. The obtained value is used as the second interaction index B.

[0098] First, the instantaneous phase time series of the respiratory signal is directly used as a low-frequency phase variable to represent the rhythmic changes of the respiratory cycle. Second, the EEG signal is bandpass filtered from 8Hz to 13Hz, and its instantaneous amplitude envelope time series is extracted using Hilbert transform as a high-frequency amplitude variable to represent the instantaneous fluctuations of local EEG energy. Then, the phase range of the respiratory signal is divided into 18 equal intervals, and the average amplitude of the EEG signal amplitude envelope within each respiratory phase interval is calculated to construct a statistical mapping between the respiratory cycle and EEG amplitude. Finally, the 18 average amplitudes are normalized to a probability distribution, and the information entropy is calculated and normalized to the natural logarithm. This calculation quantifies whether the EEG signal amplitude exhibits a regular concentrated distribution with the inspiratory or expiratory phases of the respiratory signal, thus deriving a coupling index.

[0099] This establishes a cross-frequency statistical mapping between the periodic phase of the respiratory signal and the local amplitude envelope of the EEG signal, quantifying the nonlinear modulation intensity of the respiratory cycle on EEG amplitude fluctuations. When subjects undergo deep, slow breathing intervention, the system can quantify the actual sedation or activation efficacy of the respiratory signal intervention on the EEG signal state through this significant enhancement of coupling, providing real-time quantitative closed-loop feedback for breathing training.

[0100] It can be seen that by refining the second interaction index into two orthogonal features, namely phase synchronization degree and amplitude modulation degree, the phase synchronization degree can accurately capture the dynamic alignment characteristics of the central nervous system rhythm and respiratory cycle in the time dimension to reflect the rhythmic regulation of the autonomic nervous system, and the amplitude modulation degree can quantify the nonlinear influence of respiratory activity on EEG power to reveal the feedback effect of peripheral physiological state on central cortical excitability. This effectively overcomes the limitation of single coupling index being easily affected by individual differences in respiratory frequency or environmental noise interference, enabling the system to adaptively match the neuro-respiratory response patterns of different subjects, comprehensively depict the overall physiological dynamics in the process of emotion regulation, and thus improve the robustness and stability of feedback training in complex and ever-changing practical application scenarios.

[0101] Step S404: Calculate the third interaction index between the EEG signal and the gastric electrical signal.

[0102] Among them, gastric electrical signals can be weak surface potential data reflecting the basic electrical rhythm of gastric smooth muscle and the state of autonomic nervous activity in the digestive system, and can be used to provide a deeper characterization of vagal tone and gastrointestinal motility. In this embodiment, gastric electrical signals can be acquired by collecting the superimposed slow-wave potential signals generated by gastric pacemaker cells through specific abdominal lead electrodes.

[0103] The third interaction index can be a dedicated numerical parameter that quantifies the strength of the dynamic correlation between EEG and EEG signals. It can be used to accurately map the intensity of the central nervous system's regulation of the digestive system and support the construction of multi-dimensional feedback maps. In an exemplary embodiment, the calculation of the third interaction index between EEG and EEG signals can be achieved by using wavelet transform to extract multi-scale features of EEG and EEG signals and calculating cross-entropy, and by using complex Morlet wavelet coherence analysis to evaluate the energy synchronization of the two in a specific frequency band. This allows for the independent quantification of the intensity of the central nervous system's regulation of the digestive system and improves the complete representation of the visceral autonomic neural network.

[0104] In one embodiment, the third interaction index includes a third interaction index A and a third interaction index B, wherein the third interaction index A is used to characterize the degree of signal coordination between the electroencephalogram (EEG) signal and the gastric electrical signal; and the third interaction index B is used to characterize the degree of modulation of the amplitude change of the EEG signal by the gastric electrical signal.

[0105] The third interaction index A can be a numerical parameter that quantifies the dynamic consistency between EEG oscillations and slow gastric waves in the time-frequency domain, and can be used to reflect the overall level of autonomic nervous system regulation of digestive function. In this embodiment, the third interaction index A can extract the time-frequency feature matrix of the two signals and calculate the matching degree of energy distribution or waveform morphology.

[0106] Signal coherence can be an abstract attribute describing the dynamic consistency between two physiological signals in terms of temporal evolution and frequency distribution. It can be used as an intrinsic representation dimension of index A and reflect the linkage stability between the central nervous system and peripheral organs. Furthermore, the signal coherence between EEG and Gastrointestinal signals can be characterized by using dynamic time warping algorithms to evaluate the temporal matching degree of the waveform morphology of the two signals or by using cross wavelet transform to calculate the energy overlap ratio within a specific frequency band. This allows for the quantification of the dynamic consistency between central nervous system rhythms and gastrointestinal electrical activity in the time-frequency domain.

[0107] Furthermore, the third interaction index A between the EEG signal and the gastric electrical signal is calculated, including: The EEG signal is bandpass filtered in the sixth preset frequency band and subjected to Hilbert transform to extract the instantaneous phase time series of the EEG slow wave signal; the gastric electroencephalogram (GEG) signal is subjected to Hilbert transform to extract the instantaneous phase time series of the GEG slow wave signal; the phase difference between the instantaneous phase time series of the EEG slow wave signal and the GEG slow wave signal at each time point is calculated, and the arithmetic mean of the cosine values ​​of the phase differences at all time points is obtained as the third interaction index A.

[0108] First, the EEG signal was bandpass filtered from 0.5Hz to 4Hz to specifically extract the extremely low-frequency slow-wave components, and the instantaneous phase time series of the EEG slow-wave signal was extracted using Hilbert transform. Second, the gastric electrical signal was directly subjected to Hilbert transform to extract the instantaneous phase time series of the gastric electrical slow-wave signal. Since the basic gastric electrical rhythm itself is an extremely low-frequency slow wave, its periodic phase can be directly extracted without additional filtering. Finally, the phase difference between the EEG slow wave and the gastric electrical slow wave at each time point was calculated, and the arithmetic mean of the cosine values ​​was obtained. This calculation rigorously evaluated the coherence and synchronization lock of the two slow-wave signals in the same extremely low-frequency physical dimension.

[0109] This approach, targeting the characteristics of extremely low-frequency signals, enables instantaneous phase alignment and coherence calculation of the slow-wave components of EEG and Gastrointestinal electrical signals on the same low-frequency physical dimension. Since the slow-wave components of Gastrointestinal electrical signals are highly susceptible to rhythmic disruption by external stimuli, this synchronization index can objectively quantify the degree of interference of emotional stress on the basic rhythm of Gastrointestinal electrical signals, providing a specific dimension for assessing somatization symptoms and relaxation effects.

[0110] The third interaction index B can be a numerical parameter that quantifies the periodic shaping effect of gastric electrical rhythm on brain electrical power, and can be used to characterize the bottom-up physiological information transmission mechanism of the gut-brain axis. In an exemplary embodiment, the third interaction index B can be used to extract the segmented envelope of the brain electrical signal and calculate the modulation depth based on the basic gastric electrical rhythm. Furthermore, the degree of modulation of the brain electrical signal amplitude change by the gastric electrical signal can be characterized by using a modulation index algorithm to assess the fluctuation depth of brain electrical amplitude with the gastric slow wave cycle or by using a synchronization likelihood method to analyze the periodic influence of gastric electrical phase on brain electrical power, thereby characterizing the periodic shaping effect of peripheral gastrointestinal rhythm on central nervous system excitability.

[0111] Furthermore, the third interaction index B between the electroencephalogram (EEG) signal and the gastric electrical signal is calculated, including: The instantaneous phase time series of the gastric electroencephalogram (GEG) signal is used as a low-frequency phase variable. The EEG signal is bandpass filtered in the seventh preset frequency band and subjected to Hilbert transform to extract the instantaneous amplitude envelope time series of the EEG signal, which is used as a high-frequency amplitude variable. The range of the low-frequency phase variable is divided into multiple equal phase intervals, and the average amplitude of the high-frequency amplitude variable in each phase interval is calculated. The multiple average amplitudes are normalized to obtain a probability distribution, the information entropy of the probability distribution is calculated, and the information entropy is normalized by dividing the information entropy by the natural logarithm of the maximum possible information entropy. The obtained value is used as the third interaction index B.

[0112] First, the instantaneous phase time series of the gastric electroencephalogram (GEG) signal is used as a low-frequency phase variable, representing the periodic changes in the basic electrical rhythm of the digestive tract. Second, the electroencephalogram (EEG) signal is bandpass filtered from 8Hz to 13Hz, and its instantaneous amplitude envelope time series is extracted using Hilbert transform, serving as a high-frequency amplitude variable representing the energy fluctuations in the higher frequency bands of the EEG. Then, the phase range of the GEG signal is divided into 18 intervals, and the average amplitude of the high-frequency EEG signal within each GEG phase interval is calculated, establishing a statistical mapping between very low-frequency GEG and higher-frequency EEG. Finally, the 18 average amplitudes are normalized to a probability distribution, and the information entropy is calculated and normalized to the natural logarithm. This calculation quantifies whether the high-frequency amplitude of the EEG signal is regularly modulated by the basic slow-wave phase of the GEG signal, deriving the final coupling index.

[0113] This constructs a cross-frequency modulation model between the phase of very low-frequency gastric electrical signals and the amplitude envelope of higher-frequency electroencephalogram (EEG) signals, quantifying the underlying driving strength of the gastric electrical baseline rhythm on high-frequency EEG activity. It reflects the continuous upward modulation effect of the gastric electrical signal physiological rhythm on the EEG signal state. Anomaly tracking of this indicator can be used to assess chronic fatigue or long-term gastrointestinal signal disorders, providing baseline assessment and efficacy verification for long-term neurofeedback training.

[0114] It can be seen that by refining the third interaction index into two orthogonal features—signal coordination degree and amplitude modulation degree—signal coordination degree accurately quantifies the dynamic consistency between central nervous system rhythms and gastrointestinal electrical activity in the time-frequency domain, reflecting the overall regulatory level of the autonomic nervous system on digestive function. Amplitude modulation degree, on the other hand, characterizes the nonlinear feedback effect of gastric electrical activity on brain electrical power, revealing the bottom-up physiological information transmission mechanism of the gut-brain axis. The combination of the two effectively overcomes the limitations of single indicators being susceptible to individual differences in gastrointestinal peristalsis rhythms or motion artifacts. Using this interaction index as the core regulatory target for feedback training enables the system to adaptively match the neuro-gastrointestinal response patterns of different subjects, comprehensively characterizing the overall physiological dynamics during emotion regulation, thereby improving the robustness and stability of feedback training in complex and ever-changing real-world application scenarios.

[0115] In this embodiment, peripheral physiological signals are specifically decomposed into three typical visceral signals: electrocardiogram (ECG), respiration, and gastric electroencephalogram (GEG). These signals are mapped to the autonomic nervous activity characteristics of the cardiovascular, respiratory, and digestive systems, respectively. Based on this, the coupling calculation of EEG signals with these three types of visceral signals is performed independently to generate first, second, and third interaction indicators, forming a multi-dimensional brain-visceral synergistic regulation map. This parallel computing architecture overcomes the limitations of traditional single peripheral indicators, which are easily affected by individual physiological baseline differences or environmental noise interference. It can accurately quantify the independent linkage strength of the central nervous system to different visceral organs. Since emotions and overall physiological states are essentially collaborative responses of multi-organ autonomic neural networks, the parallel calculation of multiple indicators effectively improves the robustness and comprehensiveness of feature representation. Using this multi-dimensional coupling state as the core regulatory target for feedback training enables the system to adaptively match the neurophysiological response patterns of different subjects, thereby improving the stability of feedback training.

[0116] Step S405: Map the interaction indicators to feedback parameters and present the feedback parameters to the subjects to guide them to self-regulate their physiological state based on the feedback parameters.

[0117] Example 3: The feedback training method will be described in detail below based on the mapping and presentation details of the feedback parameters.

[0118] Please see Figure 5 , Figure 5 This is a flowchart illustrating another feedback training method provided in an embodiment of this application. This method is applied to the aforementioned data processing module, such as... Figure 5 As shown, the method includes the following steps.

[0119] Step S501: Acquire the subject's electroencephalogram (EEG) signals and physiological signals.

[0120] Step S502: Calculate the interaction index between EEG signals and physiological signals.

[0121] Step S503: Based on the magnitude, degree of change, and degree of deviation of the interaction index, the feedback parameters are mapped to obtain the feedback parameters.

[0122] Among them, the degree of change is used to characterize the degree of change between the interaction indicator and the previously generated interaction indicator, the degree of deviation is used to characterize the degree of deviation of the value of the interaction indicator from the preset value range, and the feedback parameters include feedback intensity, feedback frequency, feedback rhythm, feedback continuity time or feedback level.

[0123] The preset numerical range can represent the target numerical range boundary of the ideal interaction indicator, and can be used as a reference benchmark to assess the difference between the current value of the interaction indicator and the target. In this embodiment, the preset numerical range can be set with upper and lower limit thresholds based on clinical baseline data or individualized assessment results of the subject. The degree of deviation can be a quantified value of the distance between the current value of the interaction indicator and the boundary of the target range, and can be used to represent the difference between the subject's self-regulation process and the preset target. For example, the degree of deviation can include, but is not limited to, one or more of the following: upper limit overshoot, lower limit undershoot, and offset within the range.

[0124] The degree of change can be the dynamic evolution amplitude of the interaction indicator within a continuous time window, and can be used to characterize the instantaneous change trend and adjustment response speed of the interaction indicator. In this embodiment, the degree of change can include, but is not limited to, one or more of the following: positive increment amplitude, negative decrement amplitude, and oscillation amplitude. The previously generated interaction indicator can be a snapshot of the coupling degree parameters recorded at historical sampling times, which can be used as a time-series reference benchmark for calculating the degree of dynamic change.

[0125] Feedback intensity can be the energy or amplitude parameter of the feedback parameter in the physical perception dimension, and can be used to directly stimulate the subject's senses to enhance attention regulation. In this embodiment, feedback intensity may include, but is not limited to, one or more of the following: light intensity level, sound pressure level amplitude, tactile vibration acceleration, etc.

[0126] Feedback frequency can be a periodic parameter in which feedback parameters repeat within a unit of time, and can be used to match the subject's physiological rhythm to enhance perceptual synchronicity. For example, feedback frequency can be one or more of the following: low-frequency pulse sequences, mid-frequency rhythm sequences, and high-frequency continuous sequences.

[0127] Feedback rhythm can be a pattern parameter in which feedback parameters alternate between strong and weak or are distributed intermittently over time, and can be used to guide subjects to establish specific breathing or cognitive regulation rhythms. Furthermore, feedback rhythm can include, but is not limited to, one or more of the following: uniform interval pattern, gradual acceleration pattern, random perturbation pattern, etc.

[0128] The feedback duration can be a time span parameter for the continuous output of feedback parameters, and can be used to maintain the stability and persistence of the subject's regulated state. In an exemplary embodiment, the feedback duration may include, but is not limited to, one or more of the following: instantaneous triggering period, short-term maintenance period, and long-term consolidation period.

[0129] Feedback levels can be discretized status indicators of feedback parameters across a comprehensive perception dimension, and can be used to provide simplified goal achievement cues to reduce cognitive load. In one specific embodiment, feedback levels may include, but are not limited to, one or more of the following: warning prompt levels, adjustment in progress levels, and goal achievement levels.

[0130] Furthermore, mapping interaction indicators to feedback parameters can be achieved by using a multivariate regression model to map three-dimensional features to the photoacoustic tactile parameter space, and by using a fuzzy logic rule base for condition matching and signal generation. This can achieve the technical effect of dynamically transforming abstract coupled parameters into intuitive perceptual stimuli.

[0131] In this embodiment, the system comprehensively quantifies the absolute value of the interaction index, the degree of change with the preceding index, and the degree of deviation from the preset value range. It accurately captures the instantaneous characteristics, temporal evolution patterns, and distance to the control target of the subject's brain-visceral coupling state. Based on this, the multidimensional mapping results are adaptively converted into diverse feedback parameters such as feedback intensity, frequency, rhythm, continuous time, or level. This achieves parameterization, personalization, and dynamic matching of feedback stimuli, effectively overcoming the shortcomings of traditional fixed threshold or single-dimensional feedback that are susceptible to physiological drift and noise interference. It enables subjects to make precise self-adjustments based on real-time, dynamic, and clearly goal-oriented feedback parameters, which can improve the robustness and stability of cross-individual and cross-scenario feedback training.

[0132] Step S504: Present feedback parameters to the subject based on the target form to guide the subject to self-regulate his / her physiological state according to the feedback parameters.

[0133] The target form includes one or more of visual, auditory, and tactile senses. The target form can be a set of sensory channel types that carry physical stimuli for feedback parameters, and can be used to provide a multimodal sensory carrier, enhancing the perceptual salience and anti-interference capability of feedback information. In this embodiment, the target form can be dynamically configured or preset according to the subject's sensory preferences, environmental noise levels, and training phase requirements.

[0134] Vision can be a sensory pathway that uses a display terminal or light-emitting device to convert electrical signals into visible light of a specific wavelength and intensity. This light wave stimulates the retina, triggering nerve impulses. It can be used to provide feedback stimuli with a clear spatial distribution and a large information capacity. In an exemplary embodiment, vision can include, but is not limited to, one or more of the following: monochromatic light pulses, multicolor light gradients, and dynamic graphics rendering.

[0135] Hearing can be a sensory pathway that uses loudspeakers or bone conduction devices to convert electrical signals into mechanical waves of specific frequencies and amplitudes. These sound waves vibrate the eardrum, triggering an auditory nerve response. This pathway can provide feedback stimuli with high temporal resolution and the ability to penetrate visual obstructions. Furthermore, hearing can employ pure frequency modulation, complex timbre rhythm, and spatial sound field localization techniques.

[0136] Tactile sensation can be achieved by using a vibrating motor or piezoelectric ceramic sheet to convert electrical signals into physical contact force with a specific frequency and amplitude. This force, or temperature change, stimulates skin receptors, triggering neural signals through sensory channels. It can provide feedback stimulation with low cognitive load and resistance to ambient light and sound interference. In one specific embodiment, tactile sensation can include, but is not limited to, one or more of the following: dot matrix vibration array, flexible pressure wrapping, and electrical stimulation.

[0137] Presenting feedback parameters to subjects based on the target form can involve reading the target form configuration parameters, activating the corresponding physical output actuator, converting the feedback parameter data into physical stimuli for specific sensory channels, and then outputting them.

[0138] In this embodiment, by introducing a multimodal perception channel as the presentation carrier of feedback parameters and constructing a dynamic matching mechanism based on target form, the flexible switching or fusion of visual, auditory and tactile channels effectively overcomes the defects of fatigue, attention distraction and environmental noise interference that are easily caused by single sensory feedback, and can improve the perceptual salience and cognitive accessibility of feedback information.

[0139] The following specific examples illustrate this application: S1, the subject prepares to wear the sensor.

[0140] After entering the training environment, the subject remains in a quiet sitting or semi-recumbent position. Depending on the training needs, EEG acquisition electrodes are worn on the subject's head, and ECG, respiration, and gastric electroencephalogram (GEG) sensors are deployed on the chest and / or abdomen to acquire EEG signals and visceral rhythm-related physiological signals. After the electrodes are fitted, the signal quality of each channel is checked to confirm stable electrode contact, normal sensor operation, and continuous output of valid signals from each acquisition channel.

[0141] S2, synchronous acquisition of multimodal signals.

[0142] The data acquisition module is activated to simultaneously acquire one or more of the following signals: electroencephalogram (EEG), electrocardiogram (ECG), respiratory signal, and electrogastric signal (EGG). During acquisition, all signals are synchronized and buffered under a unified time reference and transmitted in real time to the real-time analysis and interactive index calculation module.

[0143] S3, Establishing the resting baseline.

[0144] Before formal training begins, subjects undergo resting recordings for a preset duration. During this phase, the system collects multimodal physiological signals from the subjects at rest, performs preprocessing, feature extraction, and calculates brain-visceral rhythm interaction indices to establish an individual resting baseline. This individual resting baseline is used to determine the target range, feedback threshold, feedback sensitivity, and initial closed-loop control parameters for subsequent training.

[0145] S4, online preprocessing and rhythm feature extraction.

[0146] During training, the data processing module performs online preprocessing on the synchronously acquired multimodal signals, including filtering, baseline correction, artifact detection, abnormal segment processing, and rhythm extraction. Specifically, it extracts target frequency band power, instantaneous phase, and / or amplitude envelope from EEG signals; R-peaks and RR interval sequences from ECG signals; respiratory cycle, respiratory rate, respiratory amplitude, and instantaneous respiratory phase from respiratory signals; and gastric slow wave dominant frequency, dominant power, and instantaneous phase from gastric electrocardiogram (ECG) signals.

[0147] S5, Calculation of brain-visceral rhythm interaction index.

[0148] Based on the modal rhythm features extracted in step S4, the system further calculates brain-heart, brain-respiration, and brain-stomach interaction indicators. These interaction indicators may include one or more of the following: cardiac evoked potentials, EEG-HRV coupling characteristics, phase synchronization indicators, coherence indicators, phase-amplitude coupling indicators, and directional information transmission indicators. The system uses these interaction indicators to assess the subject's current brain-visceral synergistic regulatory state in real time and determines its degree of deviation and trend relative to the target state.

[0149] S6, Feedback parameter generation.

[0150] The feedback mapping unit in the data processing module converts the interaction metrics into perceptible feedback parameters. Feedback parameters may include visual feedback parameters, auditory feedback parameters, and / or tactile feedback parameters, such as image brightness, color, graphic size, pitch, volume, background music rhythm, vibration intensity, and vibration frequency.

[0151] S7, feedback presentation and autonomous adjustment.

[0152] The feedback presentation module outputs visual, auditory, and / or tactile feedback to the subject based on feedback parameters. When the brain-visceral rhythm interaction index improves towards the target range, the system outputs enhanced positive feedback; when the index deviates from the target range, the system outputs diminished or cue-like feedback. Based on the feedback information, the subject actively adjusts their own state, including breathing rhythm, relaxation level, attentional state, or emotional regulation strategies, thereby causing corresponding changes in brain activity and visceral rhythm states.

[0153] S8, closed-loop dynamic adjustment.

[0154] During the feedback output process, the closed-loop adjustment unit in the data processing module continuously receives real-time updated brain-visceral rhythm interaction indicators and their historical changes, and compares them with preset target intervals, target trends, or individual baseline states. When the interaction indicators change towards the target state, positive feedback is enhanced; when the interaction indicators deviate from the target state or fluctuate abnormally, feedback is weakened or the guidance strategy is adjusted. Thus, the feedback intensity, feedback rhythm, feedback threshold, and training strategy are dynamically updated to maintain the stability and effectiveness of the training process.

[0155] S9, Training End and Result Output.

[0156] Once the preset training duration is reached or the training termination conditions are met, the system stops providing feedback and stores the multimodal raw data, preprocessing results, changes in interactive metrics, and training results from the training process. Simultaneously, the human-computer interface outputs the training results, including the trend of target metrics, training achievement status, and interim training effectiveness.

[0157] S10, Individualized Updates and Repeated Training.

[0158] In subsequent training, the system can access historical training data to update the individual's resting baseline, target range, and feedback strategies. Through multiple training iterations, a personalized closed-loop training program suitable for the subject is gradually formed, thereby improving the continuity, stability, and effectiveness of emotion regulation training.

[0159] For embodiments consistent with those shown above, please refer to... Figure 6 , Figure 6 This is a functional unit block diagram of a feedback training device provided in an embodiment of this application. The feedback training device is the data processing module or a part of the data processing module described above, such as... Figure 6 As shown, the feedback training device 60 includes: The communication unit 601 is used to acquire the subject's electroencephalogram (EEG) signals and physiological signals. The physiological signals are used to reflect the body's physiological state, and the EEG signals and physiological signals have the same timestamp. The processing unit 602 is used to calculate the interaction index between the electroencephalogram (EEG) signal and the physiological signal. The interaction index is used to characterize the degree of coupling between the subject's EEG signal and the physiological signal. The higher the value of the interaction index, the higher the degree of coupling between the EEG signal and the physiological signal. The processing unit 602 is also used to map interaction metrics to feedback parameters; The communication unit 601 is also used to present feedback parameters to the subject in order to guide the subject to self-regulate his / her physiological state based on the feedback parameters.

[0160] In one feasible embodiment, the physiological signals include electrocardiogram (ECG) signals, respiratory signals, and gastric electrical signals; the interaction indicators include a first interaction indicator, a second interaction indicator, and a third interaction indicator; and the calculation of the interaction indicators between the electroencephalogram (EEG) signals and the physiological signals includes: Calculate the first interaction index between EEG and ECG signals; Calculate the second interaction index between electroencephalogram (EEG) signals and respiratory signals; Calculate the third interaction index between EEG signals and gastric electrical signals.

[0161] In one feasible embodiment, the first interaction indicator includes first interaction indicator A, first interaction indicator B, first interaction indicator C, first interaction indicator D, and first interaction indicator E, wherein, The first interaction index A is used to characterize the response intensity of EEG signals to periodic pulses of ECG signals; The first interaction index B is used to characterize the degree of linear coordination between EEG signals and ECG signals; The first interaction index C is used to characterize the degree of phase locking between EEG signals and ECG signals; The first interaction index D is used to characterize the information flow and causal dominance direction between the characteristic sequences of EEG and ECG signals. The first interaction index E is used to characterize the degree to which the electrocardiogram (ECG) signal modulates the amplitude changes of the electroencephalogram (EEG) signal.

[0162] In one feasible embodiment, the second interaction indicator includes second interaction indicator A and second interaction indicator B, wherein... The second interaction index A is used to characterize the degree of phase synchronization between EEG signals and respiratory signals; The second interaction index B is used to characterize the degree to which respiratory signals modulate changes in the amplitude of electroencephalogram (EEG) signals.

[0163] In one feasible embodiment, the third interaction metric includes third interaction metric A and third interaction metric B, wherein... The third interaction index A is used to characterize the degree of signal coordination between electroencephalogram (EEG) signals and gastric electrical signals. The third interaction index, B, is used to characterize the degree to which gastric electrical signals modulate the amplitude changes of electroencephalogram (EEG) signals.

[0164] In one feasible embodiment, in terms of mapping interaction metrics to feedback parameters, processing unit 602 is specifically configured to: Feedback parameters are obtained by mapping the numerical magnitude, degree of change, and degree of deviation of the interaction indicator. The degree of change is used to characterize the degree of change between the interaction indicator and the previously generated interaction indicator. The degree of deviation is used to characterize the degree of deviation of the value of the interaction indicator from the preset numerical range. Feedback parameters include feedback intensity, feedback frequency, feedback rhythm, feedback continuity time, or feedback level.

[0165] In one feasible embodiment, in presenting feedback parameters to the subject, the communication unit 601 is specifically used for: Feedback parameters are presented to subjects based on target forms, which include one or more of the visual, auditory, and tactile forms.

[0166] It is understood that since the method embodiments and the device embodiments are different presentations of the same technical concept, the content of the method embodiment section in this application should be adapted to the device embodiment section in a synchronous manner, and will not be repeated here.

[0167] When using integrated units, such as Figure 7 As shown, Figure 7 This is a block diagram of the functional units of another feedback training device provided in an embodiment of this application. Figure 7 In this embodiment, the feedback training device 60 includes a processing module 712 and a communication module 711. The processing module 712 controls and manages the actions of the feedback training device 60, such as the steps of the processing unit 602, and / or performs other processes using the techniques described herein. The communication module 711 supports interaction between the feedback training device 60 and other devices, such as the steps of the communication unit 601. Figure 7 As shown, the feedback training device 60 may also include a storage module 713, which is used to store the program code and data of the feedback training device 60.

[0168] The processing module 712 can be a processor or controller, such as a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), an ASIC, an FPGA, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor can also be a combination that implements computational functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc. The communication module 711 can be a transceiver, RF circuitry, or a communication interface, etc. The storage module 713 can be a memory.

[0169] All relevant content for each scenario involved in the above method embodiments can be referenced from the functional descriptions of the corresponding functional modules, and will not be repeated here. The above-mentioned feedback training device 60 can execute the feedback training methods shown in Embodiments 1 to 3.

[0170] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. A computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0171] Figure 8 This is a structural block diagram of an electronic device provided in an embodiment of this application. Figure 8 As shown, the electronic device 800 may include one or more of the following components: processor 801, memory 802 and communication interface 803. The processor 801, memory 802 and communication interface 803 are interconnected and perform communication between them. The memory 802 may store one or more computer programs, which may be configured to implement the methods described in the above embodiments when executed by one or more processors 801.

[0172] Processor 801 may include one or more processing cores. Processor 801 connects to various parts within the electronic device 800 using various interfaces and lines, and performs various functions and processes data of the electronic device 800 by running or executing instructions, programs, code sets, or instruction sets stored in memory 802, and by calling data stored in memory 802. Optionally, processor 801 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). Processor 801 may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. It is understood that the aforementioned modem may also not be integrated into processor 801, but may be implemented separately through a communication chip.

[0173] The memory 802 may include random access memory (RAM) or read-only memory (ROM). The memory 802 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 802 may include a program storage area and a data storage area. The program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as touch functionality, sound playback functionality, image playback functionality, etc.), and instructions for implementing the various method embodiments described above. The data storage area may also store data created by the electronic device 800 during use.

[0174] It is understood that the electronic device 800 may include more or fewer structural elements than those shown in the above block diagram, such as a power module, physical buttons, WiFi (Wireless Fidelity) module, speaker, Bluetooth module, sensor, etc., without limitation.

[0175] The aforementioned electronic device 800 may be a data processing module or a part of a data processing module.

[0176] This application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements some or all of the steps of any of the feedback training methods described in the above method embodiments.

[0177] This application also provides a computer program product, including a computer program that, when executed by a processor, implements some or all of the steps of any of the feedback training methods described in the above method embodiments. This computer program product can be a software installation package.

[0178] It should be noted that, for the sake of simplicity, each of the aforementioned feedback training method embodiments is described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to this application.

[0179] Although this application has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings, disclosure, and appended claims, will understand and implement other variations of the disclosed embodiments in carrying out the claimed application. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude multiple instances. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce a good effect.

[0180] Those skilled in the art will understand that all or part of the steps in the various method embodiments of any of the above-described feedback training methods can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage device, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a disk, or an optical disk, etc.

[0181] The embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of a feedback training method, apparatus, electronic device, and storage medium of this application. The descriptions of the embodiments above are only for the purpose of helping to understand the method and its core ideas of this application. At the same time, for those skilled in the art, based on the ideas of a feedback training method, apparatus, electronic device, and storage medium of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

[0182] This application is described with reference to flowchart illustrations and / or block diagrams of methods, hardware products, and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0183] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0184] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0185] It is understood that any product that is controlled or configured to execute the processing method of the flowchart described in the method embodiment of the feedback training method of this application, such as the terminal and computer program product of the above flowchart, falls within the scope of the related products described in this application.

[0186] Obviously, those skilled in the art can make various modifications and variations to the feedback training method, apparatus, electronic device, and storage medium provided in this application without departing from the spirit and scope of this application. Therefore, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application also intends to include these modifications and variations.

Claims

1. A feedback training method, characterized in that, The method includes: The subject's electroencephalogram (EEG) and physiological signals are acquired, wherein the physiological signals are used to reflect the body's physiological state, and the EEG and physiological signals have the same timestamp. The interaction index between the electroencephalogram (EEG) signal and the physiological signal is calculated. The interaction index is used to characterize the degree of coupling between the EEG signal and the physiological signal of the subject. The higher the value of the interaction index, the higher the degree of coupling between the EEG signal and the physiological signal. The interaction indicators are mapped to feedback parameters, and the feedback parameters are presented to the subject to guide the subject to self-regulate his / her physiological state based on the feedback parameters.

2. The method according to claim 1, characterized in that, The physiological signals include electrocardiogram (ECG) signals, respiratory signals, and gastric electrical signals. The interaction indicators include a first interaction indicator, a second interaction indicator, and a third interaction indicator. Calculating the interaction indicators between the electroencephalogram (EEG) signals and the physiological signals includes: Calculate the first interaction index between the electroencephalogram (EEG) signal and the electrocardiogram (ECG) signal; Calculate a second interaction index between the electroencephalogram (EEG) signal and the respiratory signal; Calculate the third interaction index between the electroencephalogram (EEG) signal and the gastric electrical signal.

3. The method according to claim 2, characterized in that, The first interaction indicator includes first interaction indicator A, first interaction indicator B, first interaction indicator C, first interaction indicator D, and first interaction indicator E, wherein, The first interaction index A is used to characterize the response intensity of the EEG signal to the periodic pulses of the ECG signal; The first interaction index B is used to characterize the degree of linear coordination between the electroencephalogram (EEG) signal and the electrocardiogram (ECG) signal; The first interaction index C is used to characterize the degree of phase lock between the electroencephalogram (EEG) signal and the electrocardiogram (ECG) signal; The first interaction index D is used to characterize the information flow direction and causal dominance direction between the feature sequences of the EEG signal and the ECG signal; The first interaction index E is used to characterize the degree of modulation of the amplitude change of the electrocardiogram signal by the electroencephalogram signal.

4. The method according to claim 2 or 3, characterized in that, The second interaction metric includes second interaction metric A and second interaction metric B, wherein, The second interaction index A is used to characterize the degree of phase synchronization between the electroencephalogram (EEG) signal and the respiratory signal; The second interaction index B is used to characterize the degree to which the respiratory signal modulates the amplitude change of the electroencephalogram signal.

5. The method according to any one of claims 2-4, characterized in that, The third interaction indicator includes third interaction indicator A and third interaction indicator B, wherein... The third interaction index A is used to characterize the degree of signal coordination between the electroencephalogram (EEG) signal and the gastric electroencephalogram (GEG) signal. The third interaction index B is used to characterize the degree of modulation of the amplitude change of the electroencephalogram by the gastric electroencephalogram.

6. The method according to any one of claims 1-5, characterized in that, The step of mapping the interaction metrics to feedback parameters includes: Feedback parameters are obtained by mapping the numerical magnitude, degree of change, and degree of deviation of the interaction indicator. The degree of change is used to characterize the degree of change between the interaction indicator and the previously generated interaction indicator. The degree of deviation is used to characterize the degree of deviation of the value of the interaction indicator from the preset numerical range. The feedback parameters include feedback intensity, feedback frequency, feedback rhythm, feedback continuity time, or feedback level.

7. The method according to any one of claims 1-6, characterized in that, Presenting the feedback parameters to the subject includes: The feedback parameters are presented to the subject based on a target form, which includes one or more of visual, auditory, and tactile sensations.

8. A feedback training device, characterized in that, The device includes: A communication unit is used to acquire the subject's electroencephalogram (EEG) signals and physiological signals, wherein the physiological signals are used to reflect the body's physiological state, and the EEG signals and the physiological signals have the same timestamp. The processing unit is used to calculate the interaction index between the electroencephalogram (EEG) signal and the physiological signal. The interaction index is used to characterize the degree of coupling between the EEG signal and the physiological signal of the subject. The higher the value of the interaction index, the higher the degree of coupling between the EEG signal and the physiological signal. The processing unit is also configured to map the interaction index into feedback parameters; The communication unit is also used to present the feedback parameters to the subject in order to guide the subject to self-regulate his / her physiological state based on the feedback parameters.

9. An electronic device, the device comprising a processor, a memory, and executable program code stored in the memory, characterized in that, The processor is configured to retrieve the executable program code stored in the memory to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-7.

Images