Brain age prediction method and system based on sleep electroencephalogram (EEG) signals

By using a deep learning model based on sleep EEG signals and combining it with brain structure data, a non-invasive and low-cost method for predicting individual brain age and evaluating the treatment effect of sleep disorders has been achieved. This solves the problem that existing technologies cannot balance accessibility and accuracy, and is suitable for individual brain age prediction and sleep disorder monitoring in primary healthcare institutions.

CN122208167APending Publication Date: 2026-06-16CHINA JAPAN FRIENDSHIP HOSPITAL
0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
CHINA JAPAN FRIENDSHIP HOSPITAL
Filing Date
2026-01-14
Publication Date
2026-06-16

AI Technical Summary

⚠Technical Problem

Existing technologies cannot achieve non-invasive, low-cost, repeatable, and accurate prediction of individual brain age, nor can they effectively assess the synergistic changes in brain function and structure, especially in sleep disorders where dynamic monitoring and intervention effect evaluation are lacking.

⚗Method used

Sleep EEG signals were collected using a polysomnography system. Convolutional neural networks (CNN) and bidirectional long short-term memory networks (Bi-LSTM) were combined to extract features through continuous wavelet transform (CWT), and a deep learning model was constructed. Combined with Pearson correlation analysis and multiple linear regression, brain age prediction and correlation of brain structural aging characteristics were achieved, and the treatment effect of sleep disorders was dynamically monitored.

🎯Benefits of technology

It achieves non-invasive, low-cost, and highly accurate individual brain age estimation, filling the technological gap in individual-level sleep EEG brain age prediction. It can dynamically monitor the effect of sleep disorders on brain age improvement and is suitable for primary healthcare institutions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
Patent Text Reader

Abstract

The brain age prediction method and system based on sleep electroencephalogram (EEG) signal can realize non-invasive, low-cost and high-accuracy individual brain age estimation, fill the technical gap of "individual sleep EEG brain age prediction", and solve the core contradiction that the existing technology cannot balance "popularity" and "accuracy". The method comprises the following steps: (1) collecting and preparing data; (2) preprocessing sleep EEG data; (3) extracting EEG features; (4) constructing and training a deep learning brain age prediction model; (5) correlation analysis of BAI and brain structure aging characteristics; (6) grouping the subjects according to the sleep disorder type, and statistically comparing the EEG spectral power of the target brain area of each group; (7) dynamic monitoring.
Need to check novelty before this filing date? Find Prior Art