Pre-Seizure Aura Detection Using Facial Micro Expression and Blink-Rate Analysis on Raspberry Pi

View at publisher Summarise with VIRAT
Authors: S. Anbumani, Chiranjeevi T B, Karthiyayini M, Karthikeyan S
Cited by: 0
References: 23

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Epileptic seizures are frequently preceded by subtle physiological and behavioral signals called pre-seizure auras. Detecting these early cues can provide patients and caregivers valuable lead time to reduce harm. This paper proposes a low-cost, non-invasive edge solution implemented on a Raspberry Pi that continuously monitors facial micro expressions and blink-rate dynamics. The system computes Eye Aspect Ratio (EAR) to detect blinks and extracts texture and gradient features (Local Binary Patterns and Histogram of Oriented Gradients) from facial regions to capture micromovements. A lightweight fusion classifier combines blink metrics and micro expression descriptors to issue an on-device alert (buzzer/LED) and log events locally. Early experiments and simulation demonstrate promising real-time performance for seizure types that manifest facial or ocular premonitory signs. Clinical validation is required before medical use.

Fields of Study & Subjects

Emotion and Mood Recognition EEG and Brain-Computer Interfaces Epilepsy research and treatment

Publication details

Journal / source
International Conference on Communication, Computing and Electronics Systems (ICCCES)
Pages
817–822
Publication date
21 January 2026
Language
EN