An Epilepsy Detection System based on Smartphone Acceleration Sensors

  • Dr. Saad Abdalratha Makki, Mohammed Fahmy Abdul Ghafoor
Keywords: Epilepsy classification integration normalization trapezoidal seizures


Epilepsy is a neurological disorder that might have severe consequences for people's lives. Therefore, it becomes essential to use tools for the early detection of attacks. The fast technological advances in mobile communication have contributed to building specialized tools to help the patients and doctors of early detection of epilepsy seizures. The proposed system capable of monitoring and recording the patient's movement patterns is presented in this paper. This system can differentiate between the normal movement of the patient and the one resulting from an epilepsy seizure. Gravity acceleration has been compensated and the trapezoidal integration was used for signal smoothing; these signals have been transformed into features. The normalization is then applied in order to equalize the sizes of the samples. After testing the proposed algorithm against 13 different classes, including the epilepsy class, the proposed method successfully distinguishes the abnormal (epileptic) movement from the non-epileptic ones. After conducting a series of experiments it was discovered that the proposed method can be a great tool for researchers, patients, doctors and any whom who has in touch with the neurological malady.  The the system has achieved almost 99% correlation between the input eplipsy case and the enrolled one.

How to Cite
Mohammed Fahmy Abdul Ghafoor, D. S. A. M. (2021). An Epilepsy Detection System based on Smartphone Acceleration Sensors . Design Engineering, 3747-3757. Retrieved from