Parkinson early screening method and system based on olfactory evoked brain electrical signals

By extracting multimodal features from olfactory-evoked EEG signals and modeling cross-channel feature associations, combined with attention weighting mechanisms and support vector machine classification models, we have achieved efficient and reliable early screening for Parkinson's disease, solving the problems of lag and subjectivity in traditional diagnostic methods.

CN122163140APending Publication Date: 2026-06-09XIAN UNIV OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
XIAN UNIV OF TECH
Filing Date
2026-01-30
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

Current diagnostic methods for Parkinson's disease suffer from diagnostic lag and high subjectivity, making it difficult to identify the condition early and provide an opportunity for effective intervention.

Method used

By extracting multimodal features from olfactory-evoked EEG signals, modeling cross-channel feature associations, and expanding the adaptive feature space, a dual-channel cross-dimensional feature fusion and dimensionality-upgrading framework is constructed. Combined with an attention weighting mechanism and a support vector machine classification model, early screening for Parkinson's disease is achieved.

Benefits of technology

This provides an efficient and reliable non-invasive early screening method that can more comprehensively characterize the changes in EEG signals under olfactory stimulation, improve the training efficiency of the model and the stability of classification results, and overcome the lag and subjectivity of traditional diagnosis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122163140A_ABST
    Figure CN122163140A_ABST
Patent Text Reader

Abstract

This invention discloses an early Parkinson's disease screening method based on olfactory evoked EEG signals. It constructs a dual-channel, cross-dimensional feature fusion framework through multimodal basic feature extraction, cross-channel feature association modeling, and adaptive feature space expansion. Before inputting features into the classification model, the feature vectors are standardized and normalized to eliminate the influence of differences in dimensions and numerical ranges among different features. Finally, an attention weight mechanism is introduced, and the processed feature vectors are input into a pre-trained classification model to output the classification result of early Parkinson's disease risk. This invention also discloses an early Parkinson's disease screening system based on olfactory evoked EEG signals. Through olfactory stimulation paradigms, low-guided cross-dimensional feature fusion dimensionality enhancement, and standardized data processing, this invention constructs a sensitive, objective, and efficient early Parkinson's disease screening model, effectively solving the problems of diagnostic lag, strong subjectivity, and single feature in existing technologies.
Need to check novelty before this filing date? Find Prior Art