Rapid Nasal Breathing as a Biometric Trigger: High-Accuracy Electroencephalogram-Based Authentication for Clinical...

Area of Science:

• Biometrics
• Neuroscience
• Signal Processing

Background:

• Traditional biometrics face forgery risks, necessitating advanced security.
• Electroencephalography (EEG) offers liveness detection but often needs external stimuli.
• Intrinsic respiratory signals present a novel, stimulus-free biometric approach.

Purpose of the Study:

• To investigate respiratory-evoked EEG signals for identity authentication.
• To evaluate the efficacy of different breathing patterns for biometric recognition.
• To develop a deep learning model for analyzing spatial-temporal EEG features.

Main Methods:

• A 64-channel EEG system with synchronized respiratory monitoring was used.
• Participants performed four distinct breathing patterns (oral, nasal, slow nasal, rapid nasal).
• A hybrid deep learning model processed EEG data for feature extraction.

Main Results:

• Identity recognition accuracy reached 98.3% with rapid nasal breathing-evoked EEG.
• Nasal breathing patterns demonstrated superior accuracy over oral breathing.
• Rapid nasal breathing exhibited the strongest discriminative capability.

Conclusions:

• Respiratory-evoked EEG signals serve as a robust, non-invasive biometric identifier.
• High accuracy with rapid nasal breathing supports clinical integration.
• Potential applications include continuous patient authentication and secure health record access.