IEC 2021 | Machine learning from wristband sensor data for wearable, noninvasive seizure forecasting

Christian Meisel, MD, Charité – Universitätsmedizin Berlin, Berlin Institute of Health, Berlin, Germany, describes his Epilepsia Prize-winning work on machine learning from wristband sensor data for wearable, non-invasive seizure forecasting. Dr Meisel explains the benefits of such a wrist-worn device for forecasting and describes the sensors used, before explaining how these parameters were used to train deep learning algorithms to classify seizure risk. He explains how, in collaboration with researchers at Boston Children’s Hospital, he was able to train algorithms that were able to define pre- and intra-ictal states in epileptic patients. This interview took place during the 34th International Epilepsy Congress, 2021.