Search

CN-121987914-A - Teenager concentration force closed loop system with multi-mode biological signal fusion

CN121987914ACN 121987914 ACN121987914 ACN 121987914ACN-121987914-A

Abstract

The invention discloses a multi-mode biological signal fusion teenager concentration closed loop system, and relates to the technical field of data processing. The system comprises a head-mounted acquisition device and a data processing terminal, wherein the head-mounted acquisition device is used for acquiring brain electrical signals on the surface of a head of a user, emitting near infrared light and receiving reflected light intensity to acquire cerebral hemodynamic data, acquiring IMU data of the user and transmitting audio feedback signals through a bone conduction mode, and the data processing terminal is used for calculating head movement amplitude indexes based on the IMU data, dynamically calculating reliability weight data based on the head movement amplitude indexes, dynamically weighting and fusing calculation to generate multi-mode concentration indexes, dynamically controlling the bone conduction audio transducer to output intervention signals based on the multi-mode concentration indexes, realizing accurate evaluation of high time resolution (EEG) and high spatial resolution (fNIRS), and providing non-visual-dependent immersive feedback experience.

Inventors

  • YU YUE

Assignees

  • 北京专注纪科技发展有限公司

Dates

Publication Date
20260508
Application Date
20260325

Claims (10)

  1. 1. The teenager concentration closed loop system with the multi-mode biological signal fusion is characterized by comprising a head-mounted acquisition device and a data processing terminal, wherein the head-mounted acquisition device comprises an EEG acquisition unit, a fNIRS acquisition unit, an IMU inertial measurement unit and a bone conduction audio transducer, and the data processing terminal comprises a motion artifact gating unit, a dynamic gating weighting unit, a multi-mode fusion unit and a feedback control unit: Wherein the EEG acquisition unit is used for acquiring an EEG signal of the surface of the head of the user, and the EEG signal at least comprises a signal from a forehead cortex area; The fNIRS acquisition unit is used for transmitting near infrared light and receiving reflected light intensity so as to acquire cerebral hemodynamic data; The IMU inertial measurement unit is used for collecting IMU data of a user in real time, and the IMU inertial measurement unit comprises head triaxial acceleration and angular velocity data; The bone conduction audio transducer is used for transmitting an audio feedback signal in a bone conduction mode; The motion artifact gating unit is used for calculating a head motion amplitude index based on IMU data; The dynamic gating weighting unit is used for dynamically calculating and obtaining reliability weight data based on the head movement amplitude index through a preset nonlinear confidence coefficient attenuation function, wherein the reliability weight data comprises EEG signal reliability weight and fNIRS signal reliability weight; the multi-modal fusion unit is used for extracting characteristics according to the sliding time window based on the reliability weight data, carrying out dynamic weighted fusion calculation and generating a multi-modal concentration index; the feedback control unit is used for dynamically controlling the bone conduction audio transducer to output an intervention signal based on the multi-mode concentration index.
  2. 2. The multi-modal biosignal fused adolescent concentration closed loop system of claim 1 wherein said head movement amplitude indicator is obtained as follows: ; ; ; in the formula, As an index of the head motion amplitude at the current time t, As the acceleration component at the current time t, As the angular velocity component of the current time t, And Respectively a preset acceleration component weight and an angular velocity component weight, i is the number of sampling points in the sliding window, N is the total number of the sampling points, The model value is synthesized for the triaxial acceleration, For the average value of the acceleration synthesis module value in the sliding window, The model value is synthesized for the three-axis angular velocity, The average of the modulus values is synthesized for the angular velocity within the sliding window.
  3. 3. The multi-modal biosignal fused adolescent concentration closed loop system of claim 2 wherein the confidence decay function employs a Sigmoid variant function having the mathematical expression: ; in the formula, For the EEG signal reliability weight at the current instant t, For a predetermined system sensitivity coefficient, A preset motion interference threshold value; the obtaining expression of the fNIRS signal reliability weight is as follows: ; The data processing terminal is configured to take the ratio of the theta wave to the beta wave of the EEG signal acquired by the EEG acquisition unit as a first index of concentration assessment when the head movement amplitude index is not higher than the movement interference threshold value, and take the concentration of the oxyhemoglobin in the cerebral hemodynamic data acquired by the fNIRS acquisition unit as a second index of concentration assessment when the head movement amplitude index is higher than the movement interference threshold value.
  4. 4. A multimodal biosignal fusion adolescent concentration closed loop system according to claim 3 wherein, during the initial baseline calibration phase of each wear, the data processing terminal personalizes the initial value of the motion disturbance threshold by: in an initial calibration time window after the system is started, playing a prompt tone through the bone conduction transducer to guide a user to be in a resting state; Continuously collecting IMU data in the time window to obtain a baseline motion sequence, and calculating the statistical mean value and standard deviation of the baseline motion sequence; according to a preset confidence coefficient multiplier, calculating and setting an initial value of a personalized motion interference threshold based on the statistical mean value and the standard deviation, wherein the specific expression is as follows: ; in the formula, Is the statistical mean of the baseline motion sequence, The standard deviation of the baseline motion sequence, k, is the confidence multiplier.
  5. 5. The multi-modality biosignal fused adolescent concentration closed loop system of claim 1 wherein the data processing terminal performs a low-latency data preprocessing operation based on time-domain fidelity prior to performing the multi-modality fusion unit: Acquiring a first basic filtering strength and a first adjusting unit preset by an EEG signal, and a second basic filtering strength and a second adjusting unit preset by a fNIRS signal; Determining a filtering intensity adjusting coefficient based on the head motion amplitude index output by the motion artifact gating unit; Performing product operation on the filter strength adjusting coefficient and the first adjusting unit and the second adjusting unit respectively to obtain first adjusting filter strength and second adjusting filter strength; Correcting the first basic filtering strength and the second basic filtering strength by using the first adjusting filtering strength and the second adjusting filtering strength respectively to obtain a first filtering strength corresponding to an EEG signal and a second filtering strength corresponding to a fNIRS signal; Judging whether the first filtering intensity is in a preset EEG band-pass filtering range and whether the second filtering intensity is in a preset fNIRS band-pass filtering range, if so, adopting corresponding filtering intensities to filter EEG signals and fNIRS signals respectively, and if not, adjusting the filtering intensity exceeding the range to be a boundary value of the corresponding band-pass filtering range and filtering based on the adjusted filtering intensity.
  6. 6. The multi-modal biological signal fused adolescent concentration closed loop system of claim 1 wherein the multi-modal concentration index is dynamically calculated based on confidence weight data, a first indicator of concentration assessment and a second indicator of concentration assessment, specifically: calculating the reciprocal of a first index of concentration assessment as a concentration forward component; calculating a first derivative of a second index of the concentration assessment as a blood oxygen change slope: Based on the reliability weight data, the concentration force forward component and the blood oxygen change slope, calculating to obtain a multi-mode concentration force index, wherein the multi-mode concentration force index specifically comprises the following components: ; in the formula, For the multi-modal concentration index at the current time t, In order to concentrate on the positive component of the force, To be the slope of the change in blood oxygen at time t-delta, Is a preset neurovascular coupling delay parameter.
  7. 7. The multi-modal biosignal fused adolescent concentration closed loop system of claim 6 wherein said head-mounted acquisition device further incorporates a miniature tactile motor disposed on a contact surface of said head-mounted acquisition device; The closed-loop feedback control unit is further configured to control the haptic motor to generate a micro-vibration prompt for performing attention deviation correction on the premise of not occupying visual resources of a user when the multi-mode concentration index continuously preset duration is lower than a preset multi-mode concentration threshold in the concentration training mode.
  8. 8. The multi-modal biosignal fused adolescent concentration closed-loop system of claim 1 wherein the data processing terminal is further configured to facilitate sleep and relaxation modes in which the closed-loop feedback control unit is configured to perform a breathing beat control procedure based on real-time neural feedback, comprising the steps of: after the sleep-aiding mode is started, continuously playing pink noise through the bone conduction audio transducer; comprehensively evaluating the physiological pressure index of the user based on the frequency band energy ratio of the EEG signal and the blood oxygen change slope corresponding to fNIRS signals; Determining the initial frequency of the breathing guiding beat according to the physiological pressure index, and playing breathing guiding sound corresponding to the initial frequency through the bone conduction audio transducer; after maintaining the frequency of the current respiration guidance beat to be straightened for a preset time, monitoring the change slope of the concentration of the oxyhemoglobin corresponding to the fNIRS signal in real time, and dynamically adjusting the frequency of the respiration guidance beat according to the value of the change slope of the concentration of the oxyhemoglobin; judging whether the adjusted breathing guidance beat frequency reaches a preset target resonance frequency or not; if the target resonant frequency is not reached, returning to execute the breathing guiding sound with the corresponding playing frequency; And if the target resonant frequency is reached, maintaining the target resonant frequency to play until the sleep-aiding mode is ended.
  9. 9. The multi-modal biosignal fused adolescent concentration closed loop system of claim 8 wherein the step of determining an initial frequency of respiratory guidance beats from the physiological pressure index comprises: Comparing the physiological pressure index with a preset pressure threshold; If the physiological pressure index is higher than a preset pressure threshold, determining a preset first breathing guiding beat as an initial frequency of the breathing guiding beat; And if the physiological pressure index is not higher than a preset pressure threshold, determining a preset second breathing guiding beat as an initial frequency of the breathing guiding beat.
  10. 10. The multi-modal biosignal fused adolescent concentration closed loop system of claim 8, wherein the step of dynamically adjusting the frequency of respiratory guidance beats comprises: If the gradient of the oxygen-containing hemoglobin concentration change is smaller than 0, reducing the frequency of the breathing guidance beats by a preset frequency step; And if the gradient of the oxygen-containing hemoglobin concentration change is greater than or equal to 0, keeping the frequency of the current respiration guidance beat unchanged.

Description

Teenager concentration force closed loop system with multi-mode biological signal fusion Technical Field The invention relates to the technical field of data processing, in particular to a multi-mode biological signal fusion teenager concentration closed loop system. Background During adolescence, concentration and the key stage of cognitive function development are carried out. Currently, concentration assessment and training for teenagers mainly rely on a single Electroencephalogram (EEG) technique. The intelligent brain electrical signal control system and method for concentrating force improvement disclosed in the CN118331423A patent application comprise a head-mounted image acquisition module, a head-mounted brain electrical acquisition module, a desktop eyeball tracking module and a single person data analysis module, wherein the head-mounted image acquisition module is used for being fixedly worn on the head of a user and acquiring environment image information within a preset angle range in real time, the head-mounted brain electrical acquisition module is used for being fixedly worn on the head of the user and acquiring brain electrical information of the user in real time, the desktop eyeball tracking module is used for being fixedly arranged in front of the head of the user and acquiring eye image information of the user in real time, and the single person data analysis module is used for receiving the environment image information, the brain electrical information and the eye image information, drawing a curve of brain electrical information changing along with time to obtain a brain electrical curve, and then passing through a first preset time. However, existing single EEG devices have the obvious limitation that, first, EEG signals are extremely weak (microvolts) and very sensitive to motion artifacts. Teenagers often follow movements such as head shaking, blinking or biting during use, and the electromyographic signals seriously mask real nerve signals, so that evaluation results are distorted. On the other hand, the functional Near infrared spectrum technology (fNIRS, functional Near-Infrared Spectroscopy) has better anti-interference performance on motion artifacts and high spatial resolution, but has physiological delay of a few seconds based on a hemodynamic response mechanism, and is difficult to meet the requirement of real-time feedback training on time precision. In addition, conventional feedback approaches rely mostly on visual (screen animation) or auditory (in-ear headphones). Visual feedback is easy to cause secondary distraction, and the in-ear earphone can damage teenager hearing after being worn for a long time, and cannot sense environmental sound, so that potential safety hazards exist. The prior art lacks an integrated system that combines the advantages of multiple modes, effectively resists motion disturbances, and provides safe closed loop feedback. Disclosure of Invention The embodiment of the application solves the problems of poor anti-interference capability, single feedback mechanism and hysteresis in a dynamic scene in the prior art by providing the multi-mode biological signal fusion teenager concentration closed-loop system, and realizes real-time, accurate and multi-mode closed-loop nerve feedback training on concentration state in the natural activity state of the teenager head. The embodiment of the application provides a multi-mode biological signal fusion teenager concentration closed loop system, which comprises: one or more technical solutions provided in the embodiments of the present application at least have the following technical effects or advantages: 1. The multi-mode biological signal acquisition module is integrated, motion artifact gating and dynamic nonlinear confidence weighting are adopted, time domain fidelity low-delay preprocessing, fNIRS gradient enhancement and time sequence dynamic alignment are performed, personalized baseline calibration and non-visual closed loop feedback control are performed, so that the defects that single-mode signals are easy to be interfered by motion and have physiological delay are overcome, the physiological and behavior differences of teenagers are accurately adapted, the anti-interference performance, the instantaneity and the accuracy of concentration assessment are improved, the safe and undisturbed concentration correction and relaxation sleep-aiding intervention are realized, and the integrated high-efficiency intervention targets of teenagers concentration high-precision monitoring, self-adaptive fusion assessment, full-period closed loop training and relaxation sleep-aiding are further achieved. 2. The multi-mode concentration force index is calculated based on the dynamic reliability weight fusion EEG concentration force forward component and fNIRS blood oxygen change slope, the neurovascular coupling delay parameter is dynamically and iteratively corrected to achieve multi-mode signal accurate time