Abstract:
Epilepsy is a neurological disorder which can be characterized by recurrent unprovoked
seizures caused by abnormal electrical activity in brain. Epileptic seizures have various
classifications based on their nature and impact. GTC’s i.e., Generalized Tonic Clonic
seizures are one of the most common types of seizure with a very high occurrence rate
among epileptics. GTC’s involve severe muscle contractions along with repetitive jerks
in upper as well as lower body. Keeping a track of patient’s muscle activity along with
the motion of arms in 3D space can aid in the timely detection of seizure attack and thus
can reduce the risk of severe injuries as well as SUDEP i.e., Sudden Unexpected Death
in Epilepsy. In this project a wearable band has been developed to efficiently detect
GTC’s based on a multimodal approach involving sEMG and inertial measurement unit
for monitoring muscle contractions and movement in 3D space along with the use of
heart rate fluctuations as a validating factor for seizure attack. The algorithm has been
designed by acquiring real time signals from sensors, preprocessing the signal data to
remove noise artifacts and by extracting distinguishing features that highlight the
variations in the signals while seizure occurrence. An assistive mobile application has
also been developed to be paired with wearable band and it would provide certain
features including sending an alert to the caregivers in case any seizure attack occurs
