Seizario: Novel Mobile Algorithms for Seizure and Fall Detection

Epileptic seizures and dangerous falls affect over 100 million people worldwide, including patients and seniors. There is currently no affordable solution or product for the detection of seizures and falls to assist in improving the lives of millions. Such detection problem poses several major challenges, including continuous sensing with minimal user disruption, accurate real-time emergency classification and alert notification, energy-efficient sensing and detection, and scalability and affordability of the devised solution. This paper presents Seizario; the first mobile app for continuous, energy-efficient, detection of seizures and falls to provide ambient assisted living for people with Epilepsy and fall risk. Seizario uses (only) smartphones to offer an affordable scalable solution with two main features; automatic detection of seizure convulsions and fall emergency scenarios, and immediate communication of critical information to emergency contacts. Seizario introduces novel accelerometer-based learning algorithms with elaborate finite-state-machines to automatically detect grand mal seizures and harmful falls. Seizario also offers a suite of useful features to assist in providing help for the affected population. Seizario´s detection algorithms are designed to efficiently and effectively work on mobile phones, and have shown great promise with over 95% detection rate for both seizures and falls, with minimal number of false alarms.