Seizure forecasting and detection with wearable devices are feasible

Wearable watches and a subcutaneous EEG device can independently and accurately forecast seizures in epilepsy patients. This is an important finding, as epilepsy patients consistently rate the unpredictability of seizures as one of the most limiting aspects of epilepsy.

Dr Benjamin Brinkmann (Mayo Clinic, MI, USA) and his consortium investigated various technological approaches to forecast seizures, including wearables, smartphone apps, and minimally-invasive devices [1]. Here, he allowed himself a joke: “Initially, we considered leveraging the popularity of seizure alert dogs in a wearable system, but we quickly discarded this idea out of concerns for chronic shoulder pain, as well as a possible high false-alarm rate due to cats, mail carriers, and tennis balls.” Dr Brinkmann's consortium opted instead to evaluate 2 wearable biosensors, the wrist-worn Empatica E4 and Fitbit Inspire. When discharged from the hospital, study participants received both devices, to use one while charging the other. Participants were given 1 of 3 ambulatory EEG monitoring devices for seizure confirmation: UNEEG, SubQ, or EpiMinder. Wearable and EEG data from enrolled participants were recorded for 8 months or more. Self-reported electronic seizure diaries and periodic mood and symptom surveys were recorded by participants as well.

Of the 39 included participants, only 5 discontinued during follow-up. Over 17,000 days (approx. 46 years) were recorded with over 1,700 seizures. Using the Empatica E4 device for at least 6 months, seizure forecasting was better than chance in 5 of 6 so far studied participants, with a mean AUC (area under the receiver operating characteristics curve) of 0.80. Dr Brinkmann: “On average, we predicted 2/3 of their seizures, with about 30 minutes of pre-seizure warning, and several false alarms.” He added that a data science contest on the open platform https://eval.ai replicated these outcomes. “But we do think more progress is needed to improve performance.” In 5 of 6 patients that could be studied, the sub-scalp EEG signals also independently helped to forecast seizures [2].

Using an additional cohort of 65 app users, Brinkmann's group looked for possible premonitory seizure symptoms in the next 24 hours. “Interestingly, we found a strong relation between impending self-reported seizures and patients' perceived risk as well as recently reported seizures.” Mood, sleep quality, and recent seizures were significantly predictive of seizures. These data are expected and were not yet published at the time of the AAN meeting.

1. Brinkmann B. Seizure forecasting and detection with wearable devices and subcutaneous EEG – outcomes from the My Seizure Gauge trial. PL4.001, AAN 2023 Annual Meeting, 22–27 April, Boston, USA.
2. Viana PF, et al. Epilepsia, 2022;April 8. DOI: 10.1111/epi.17252.

Copyright ©2023 Medicom Medical Publishers



Posted on 26 June 202317 May 2024