A personalized binaural beat music generation method based on electroencephalogram frequency band quantitative analysis

Abstract

This invention discloses a personalized binaural beat music generation method based on quantitative analysis of EEG frequency bands. This method automatically parses eight mainstream EEG data formats and standardizes the data through adaptive preprocessing. It uses the Welch periodogram method to calculate the power spectrum of five frequency bands, extracts key clinical indicators, and outputs a quantitative state score. Based on the EEG state, it intelligently recommends or allows manual selection of target beat frequencies, embedding the binaural beats into user-selected background audio using AM double-sideband modulation technology. A quantitative verification step for the modulation effect is introduced to ensure the effectiveness of neural feedback. This invention achieves full-process integration of EEG analysis, personalized recommendation, audio synthesis, and playback control, providing a scientific, convenient, and quantifiable technical solution for mood regulation, concentration improvement, and sleep enhancement.

Description

Technical Field

[0002] This invention relates to the fields of electroencephalogram (EEG) signal processing, digital audio synthesis, and neurofeedback technology, and in particular to a personalized binaural beat music generation method based on quantitative analysis of EEG frequency bands, which can be applied to mental health and neurorehabilitation scenarios such as emotion regulation, cognitive enhancement, and sleep intervention. Background Technology

[0004] The binaural beat effect refers to the physiological phenomenon where, when each ear receives sinusoidal sound waves with slightly different frequencies, the auditory cortex of the brain integrates these signals to generate low-frequency neural oscillations equal to the frequency difference. Current research confirms that binaural beats at specific frequencies can induce synchronization of brain electrical activity, thereby regulating cognitive and emotional states: 13-30Hz beta waves can enhance attention and alertness, 8-13Hz alpha waves can induce relaxation, and 0.5-4Hz delta waves can promote deep sleep.

[0005] Currently, mainstream binaural beat music generation tools have the following technical shortcomings:

[0006] (1) Lack of objective physiological driving mechanism: The recommendation logic relies on the user's subjective choice and cannot match the optimal beat frequency according to the individual's real-time EEG state, resulting in unstable neural feedback effect and large individual differences.

[0007] (2) Poor data compatibility: It only supports a single or a few EEG device formats, lacks automatic unit standardization and preprocessing pipeline, and makes it difficult for non-professional users to complete data import and cleaning.

[0008] (3) Fragmented functions: It does not achieve one-stop integration of EEG analysis, personalized recommendation, audio synthesis and playback control. Users need to switch between multiple tools, and the operation process is cumbersome.

[0009] (4) Lack of effect verification mechanism: It is impossible to quantitatively verify whether the binaural beats are successfully embedded in the audio, making it difficult to guarantee the effectiveness of neural feedback.

[0010] Therefore, developing a method that can automatically parse multi-format EEG data, make accurate recommendations based on objective physiological indicators, and integrate a complete audio generation and verification process has important clinical and practical value. Summary of the Invention

[0012] The purpose of this invention is to overcome the shortcomings of the prior art and provide a personalized binaural beat music generation method based on quantitative analysis of EEG frequency bands. Through the integration of the entire process of automatic parsing of multi-format EEG data, standardized preprocessing, quantitative EEG state assessment, intelligent frequency recommendation, AM modulation audio synthesis and validity verification, personalized neural music generation based on objective physiological signals is realized.

[0013] To achieve the above objectives, the present invention adopts the following technical solution:

[0014] A personalized binaural beat music generation method based on quantitative analysis of brainwave frequency bands includes the following steps:

[0015] Multi-format EEG data acquisition: Receives EEG data files input by the user, automatically identifies and parses 8 mainstream formats including .set, .edf, .fif, .bdf, .vhdr, .npy, .txt, and .csv, and supports automatic switching between reading raw continuous data and segmented event data;

[0016] Adaptive normalization preprocessing: Automatic unit detection: By calculating the peak-to-peak value and root mean square (RMS) of the original signal, the data unit is determined to be V, mV, or μV, and then multiplied by the corresponding conversion factor (V→10). 6 Amplitude normalization is achieved by (mV→10³, μV→1); Baseline correction: DC component of the signal is removed; Filtering: 0.5-50Hz bandpass filtering is performed using a 4th-order Butterworth filter, and a 50Hz IIR notch filter is used to remove power frequency interference; When the relative power of the Beta band exceeds 3 times that of the Alpha band, a 25Hz low-pass filter is automatically activated to suppress abnormal high-frequency noise; Artifact removal: Extreme artifacts such as electrooculograms and electromyograms are removed using a ±3 standard deviation criterion.

[0017] Quantitative analysis of EEG state: Power spectral density (PSD) was calculated using the Welch periodogram method with the following parameters: window length = sampling rate × 4s, overlap rate = 50%, window function = Hanning window, achieving a frequency resolution of 0.25Hz. Trapezoidal integrals were performed on five frequency bands: Delta (0.5-4Hz), Theta (4-8Hz), Alpha (8-13Hz), Beta (13-30Hz), and Gamma (30-50Hz) to obtain the absolute and relative power of each band. Key clinical indicators were extracted: The Theta / Beta ratio (TBR) was calculated as an attention assessment indicator, and the peak frequency (APF) was extracted using a peak detection algorithm as a cognitive function assessment indicator. State scoring: Based on a preset segmented mapping rule, attention scores (0-100 points), cognitive function scores (0-100 points), and relaxation state scores (0-100 points) were output, and a standardized state diagnostic report was generated.

[0018] Dual-mode target frequency determination: Automatic recommendation mode: Based on EEG analysis results, when TBR>2.0, 16Hz β wave (focus mode) is recommended; when Alpha relative power<20%, 10Hz α wave (relaxation mode) is recommended; when Delta relative power>40%, 4Hz δ wave (sleep aid mode) is recommended.

[0019] Manual selection mode: Supports users to directly specify the target frequency (4Hz / 10Hz / 16Hz).

[0020] AM Modulation Binaural Beat Synthesis: Loads user-selected background audio, automatically converts it to a 44100Hz, 16-bit, stereo standard format, and automatically copies mono audio to stereo; generates modulation signal: modulator = sin(2π×f_beat×t)×depth, where f_beat is the target beat frequency and depth is the modulation depth (default 0.15, adjustable range 0.05-0.30); stereo modulation: left channel = original left channel × (1+modulator), right channel = original right channel × (1-modulator); post-processing: applies 3-second linear fade-in and fade-out, normalizes audio peaks to ±0.9 to prevent clipping distortion;

[0021] Quantitative verification of modulation effect: Generate a 10-second pure reference modulation signal, perform Welch spectrum analysis on the differential envelope of the left and right channels of the synthesized audio, detect the peak value within ±2Hz of the target frequency, and determine the validity if the peak signal-to-noise ratio is > 20dB; Output and playback control: Save the synthesized audio as a lossless WAV format, and provide asynchronous playback, real-time progress display, linear volume adjustment, and pause / continue / stop control functions. Attached Figure Description Figure 1 This is a flowchart of a personalized binaural beat music generation method based on quantitative analysis of EEG frequency bands.

Claims

1. A personalized binaural beat music generation method based on quantitative analysis of brainwave frequency bands, characterized in that, Includes the following steps: 1) Multi-format EEG data acquisition: Automatically identifies and parses EEG data files in .set, .edf, .fif, .bdf, .vhdr, .npy, .txt, and .csv formats; 2) Adaptive standardization preprocessing: Automatically detects data units and performs amplitude standardization, sequentially performing DC removal, bandpass filtering, notch filtering, and artifact removal; 3) Quantitative analysis of EEG state: Calculates the five-band power spectrum using the Welch periodogram method, extracts the Theta / Beta ratio and Alpha peak frequency, and outputs three quantitative scores and state diagnosis; 4) Dual-mode target frequency determination: Automatically recommends target beat frequencies based on EEG analysis results, or allows users to manually select them; 5) AM Modulation Binaural Beat Synthesis: The target frequency modulation signal is superimposed onto the left and right channels of the user-selected background audio to generate stereo audio; 6) Quantitative Verification of Modulation Effect: The target frequency components are detected through spectrum analysis to verify the effectiveness of binaural beat embedding; 7) Output and Playback Control: Saved in WAV format and provides asynchronous playback and volume adjustment functions.

2. The method according to claim 1, characterized in that, The automatic unit detection described in step 2) is achieved by calculating the peak-to-peak value and root mean square value of the original signal, with a conversion coefficient of V→10. 6 mV→10³, μV→1; The filtering process uses a 4th-order Butterworth 0.5-50Hz bandpass filter and a 50Hz IIR notch filter. When the Beta power exceeds 3 times the Alpha power, the 25Hz low-pass filter is automatically enabled.

3. The method according to claim 1, characterized in that, The Welch periodogram method parameters mentioned in step 3) are window length = sampling rate × 4s, overlap rate = 50%, and Hanning window; the three scores are respectively based on the Theta / Beta ratio, Alpha peak frequency and Alpha relative power for segmented mapping.

4. The method according to claim 1, characterized in that, The AM modulation formula mentioned in step 5) is: left channel = original left channel × (1 + sin (2π × f_beat × t) × depth), right channel = original right channel × (1 - sin(2π × f_beat × t) × depth), and the modulation depth can be adjusted in the range of 0.05-0.

30.

5. The method according to claim 1, characterized in that, The modulation effect verification described in step 6) is achieved by calculating the power spectrum of the differential envelope of the left and right channels and detecting the peak signal-to-noise ratio within the target frequency ±2Hz range. When the signal-to-noise ratio is >20dB, it is considered valid.

Images