Abteilung Prächirurgische Epilepsiediagnostik – EpilepsiezentrumKlinik für Neurochirurgie

Mobile Devices for seizure detection and prediction

Patient-based seizure documentation is the standard for the steering of treatments and for many medicolegal aspects of epilepsies, yet suffers from problems in memory encoding during seizures and recall following seizures. Long-term video-EEG recordings are the gold standard for quantitative and qualitative seizure documentation, yet is limited due to needed in personal and apparative resources and by impairing patient mobility. Wearables offer new options to document seizure manifestations using multimodal assessment of motor phenomena (using accelerometry and EMG) and autonomic phenomena (pulse oxymetry, respiration, skin conductance).

The Freiburg Epilepsy Center is involved in international collaborative projects assessing detection of different seizure types using wearable devices (EU project RADAR-CNS: Remote Assessment of Disease and Relapse in collaboration with King´s College, London, UK, and in the Epilepsy Foundation project MySeizure Gauge in collaboration with Mayo Clinic, Rochester, USA). Time series of multimodal data are analyzed applying modern machine learning techniques and deep learning. Ongoing studies are presently performed in in-patients and will be transferred to an outpatient setting in 2019.