Man and the machine rise to the spike‐wave. Commentary on An automated machine learning‐based detection algorithm for spike‐wave discharges (SWDs) in a mouse model of absence epilepsy.

摘要

Anyone who has spent a considerable amount of time in visual analysis and manual scoring of EEG events of either human or animal long‐term EEG recordings fully understands that the process is not entirely a labor of love. Manual review of large volumes of EEG data, whether for detection of ictal discharges, epileptiform activity, or abnormal slowing, is notoriously time‐consuming, fatigue‐inducing, and frequently inexact. Numerous analytical software programs have been developed over the last several decades to aid visual review of continuous EEG recordings and have contributed significantly to the precision and accuracy of event detection. However, event detection by even the best algorithms remains an imperfect science. The ongoing need to build better analytical “mousetraps” for EEG event capture continues to be a daunting challenge for epileptologists, computer scientists, and basic researchers.