کنفرانس مهندسی برق ایران

صفحه اصلی / سی و چهارمین کنفرانس بین المللی مهندسی برق

Unconsciousness Classification Using Discriminative Dictionary Pair Learning

نویسندگان :

Samaneh Esmaeili¹ Hamid Behroozi² Babak Hossein Khalaj³

1- دانشگاه صنعتی شریف 2- دانشگاه صنعتی شریف 3- دانشگاه صنعتی شریف

کلمات کلیدی :

Depth of Anesthesia Monitoring،Projective Dictionary Pair Learning،Shared Dictionary Learning،EEG Spectral Pattern

چکیده :

Developing accurate and reliable methods to monitor unconsciousness during general anesthesia is essential for optimizing anesthetic administration and ensuring patient safety. This study investigates the potential of electroencephalography (EEG) patterns to develop a classification model for real-time monitoring of consciousness levels during general anesthesia. We introduce a novel discriminative dictionary pair learning model that integrates shared features and a robust classifier within the traditional Projective Dictionary Pair Learning (PDPL) framework to distinguish between conscious and unconscious states. Using 10-second EEG segments with an 8-second overlap from seven healthy volunteers who received progressively increasing propofol infusions, we showcase the model's outstanding performance in tracking the consciousness state. The high accuracy of the results indicates that EEG spectral patterns can effectively detect unconsciousness, underscoring the potential of the proposed approach for real-time monitoring in clinical anesthesia settings.