Epilepsy Database Collection Implementing Smartphone Inertial Sensors

Epilepsy is a neurological disorder characterized by recurrent seizures and these seizures require ongoing monitoring. Data collection is a crucial first step to improve our understanding of this condition. Implementing smartphone inertial sensors is a new approach to epilepsy data collection that involves using the accelerometer and gyroscope sensors in a smartphone to record physical movements during a seizure episode. The aim of data collection is to gain a deeper insight into the characteristic patterns of seizures and to improve its management. Because there was not any available dataset for epilepsy detection and all the researchers in the previous studies used locally produced datasets so, an Android application was developed that can collect raw data from motion sensors (accelerometer and gyroscope) and experiments were conducted to determine the best sample rate for epilepsy cases. In this paper, we presented an application for gathering a dataset of sensor data from 30 people 15 male and 15 females between the age of 18-45, and each participant were asked to perform 9 activities. The sampling rate used in collecting data has been investigated and analyzed using Fast Fourier Transform technique resulting a suitable sampling rate (50 samples/second) for capturing motion sensor data. The results show that smartphones can accurately record physical movements during seizures and can efficiently be utilized to collect a sizable and diverse dataset for enhancing seizure detection algorithms.