Automated seizure detection using wearable devices: A clinical practice guideline of the International League Against Epilepsy and the International Federation of Clinical Neurophysiology

Abstract

The objective of this clinical practice guideline (CPG) is to provide recommendations for healthcare personnel working with patients with epilepsy on the use of wearable devices for automated seizure detection in patients with epilepsy, in outpatient, ambulatory settings. The Working Group of the International League Against Epilepsy (ILAE) and the International Federation of Clinical Neurophysiology (IFCN) developed the CPG according to the methodology proposed by the ILAE Epilepsy Guidelines Working Group. We reviewed the published evidence using The Preferred Reporting Items for Systematic Review and Meta‐Analysis (PRISMA) statement and evaluated the evidence and formulated the recommendations following the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system. We found high level of evidence for the accuracy of automated detection of generalized tonic‐clonic seizures (GTCS) and focal‐to‐bilateral tonic‐clonic seizures (FBTCS) and recommend the use of wearable automated seizure detection devices for selected patients when accurate detection of GTCS and FBTCS is recommended as a clinical adjunct. We also found a moderate level of evidence for seizure types without GTCS or FBTCS. However, it was uncertain whether the detected alarms resulted in meaningful clinical outcomes for the patients. We recommend using clinically validated devices for automated detection of GTCS and FBTCS, especially in unsupervised patients, where alarms can result in rapid intervention (weak/conditional recommendation). At present, we do not recommend clinical use of the currently available devices for other seizure types (weak/conditional recommendation). Further research and development are needed to improve the performance of automated seizure detection and to document their accuracy and clinical utility.

Prospective Study of a Multimodal Convulsive Seizure Detection Wearable System on Pediatric and Adult Patients in the Epilepsy Monitoring Unit

“…This study may qualify for the recently proposed label of a phase III validation study ( 28 , 29 ): Multiple EMU centers were involved; the reference standard was v-EEG recordings interpreted by experts; more than 20 patients ( n = 36) with seizures were included with more than 30 seizures ( n = 66); the data and patients analyzed were disjointed from those used to develop the Detection Algorithm, removing the risk of overfitting, and all of the analyses were performed in a real-time manner fully mimicking the functioning of the algorithm on-board. Offline analysis of bio-signals may raise the possibility of overfitting to the recorded data set and can call the generalizability of results into question ( 28 ).…”

“…This study may qualify for the recently proposed label of a phase III validation study ( 28 , 29 ): Multiple EMU centers were involved; the reference standard was v-EEG recordings interpreted by experts; more than 20 patients ( n = 36) with seizures were included with more than 30 seizures ( n = 66); the data and patients analyzed were disjointed from those used to develop the Detection Algorithm, removing the risk of overfitting, and all of the analyses were performed in a real-time manner fully mimicking the functioning of the algorithm on-board. Offline analysis of bio-signals may raise the possibility of overfitting to the recorded data set and can call the generalizability of results into question ( 28 ). However, given that the Detection Algorithm was trained on separate data and a fully separate patient group, and that it was “fixed-and-frozen” before being applied to the test set, and still uses the same code that runs on-board the Embrace device, we believe that overfitting is not affecting these results, as also supported by FDA's careful evaluation.…”

“…A recent study of 255 SUDEP cases (definite and probable) and 1,148 matched controls showed that 69% of SUDEP cases in patients with GTC seizures who live alone may be prevented if patients are attended, or if their GTC seizures are controlled. Recent practical clinical guidelines recommend using clinically validated devices for automated detection of CS, the seizure types included in this study, especially in unsupervised patients, where alarms can facilitate rapid interventions ( 28 ).…”

“…Growing efforts have been made by the scientific and clinical community to standardize the studies on wearable seizure detection devices to perform a rigorous validation and to enable their ubiquitous adoption in outpatient settings ( 24 , 28 , 29 ). Published guidelines have tried to adapt the STARD criteria to the specific use case of seizure detection ( 30 ).…”

Forschung und Entwicklung telemedizinischer Anwendungen und mobiler Gesundheitstechnologien bei Epilepsie

“…Telemedicine • Wearable • Automated seizure detection • App • Digitalization sichtsarbeiten [13,15] dargestellt, liegen die Schwerpunkte mobiler Gesundheitstechnologien in der Epilepsieversorgung bei digitalen Anfallskalendern und automatisierter Anfallserkennung, dies zeigen auch die Ergebnisse unserer Umfrage. Für die Nutzung klinisch geprüfter Wearables zur automatisierten Detektion tonisch-klonischer Anfälle liegen mittleweile auch Stellungnahmen internationaler Fachgesellschaften vor [3]. Neben den "klassisch epileptologischen" Anwendungen Anfallskalender und Anfallsdetektion liefert die Umfrage Hinweise dafür, dass insbesondere für psychotherapeutische oder neuropsychologische Interventionen spezielle digitale Anwendungen genutzt werden.…”