Automated detection of nocturnal motor seizures using an audio-video system
Armand Larsen S, Terney D, Østerkjerhuus T, Vinding Merinder T, Annala K, Knight A, Beniczky S.
Brain Behav. 2022 Aug 8:e2737.
https://doi.org/10.1002/brb3.2737.
Hot off the press - Patients with nocturnal generalized tonic-clonic seizures (#TCSs), that do not share a bedroom with someone else, suffer a 67-fold increased risk of sudden unexpected death in epilepsy (#SUDEP) (Ref.1). Thus, it is crucial to develop clinically validated real-time seizure monitoring devices for patients with TCSs.
Larsen et al. (Jul. 2022) conducted a Phase-3 clinical validation study to examine the #feasibility and #accuracy of an audio-video system (#Nelli®) for real-time detection of major nocturnal motor seizures.
The study recruited 191 patients (mean age: 20 years, range: 1-72 years) and monitored their nocturnal seizures using the Nelli® system for 4183 h in total. The seizure-related results were compared with the results from electroencephalography (#EEG) recordings, the golden standard for monitoring seizures.
The system showed a high #sensitivity of 93.7% (15/16, 95% CI: 69.8%–99.8%) for monitoring major motor seizures, while a relatively low sensitivity of 8.3% (28/336, 95% CI: 5.6%–11.8%) was determined for the detection of minor motor seizures. The false detection rate was 0.16 per hour which is too high for home monitoring but is feasible for the hospital setting where medical personnel is available at all times.
The Nelli® system, with its nonobtrusive and noncontact features, might help solve the current problem in many hospitals (where the number of patients limits the possibility of continuous human/visual surveillance) by enabling high sensitivity automated detection of major motor seizures.
Ref. 1. doi: 10.1212/WNL.0000000000008741.
