Development of a Multivariable Seizure Likelihood Assessment Based on Clinical Information and Short Autonomic Activity Recordings for Children With Epilepsy - 04/10/23
, Sarah Cantley, BA a, Michele Jackson, BA a, Bo Zhang, PhD c, William J. Bosl, PhD d, e, 1, Tobias Loddenkemper, MD a, 1Abstract |
Background |
Predicting seizure likelihood for the following day would enable clinicians to extend or potentially schedule video-electroencephalography (EEG) monitoring when seizure risk is high. Combining standardized clinical data with short-term recordings of wearables to predict seizure likelihood could have high practical relevance as wearable data is easy and fast to collect. As a first step toward seizure forecasting, we classified patients based on whether they had seizures or not during the following recording.
Methods |
Pediatric patients admitted to the epilepsy monitoring unit wore a wearable that recorded the heart rate (HR), heart rate variability (HRV), electrodermal activity (EDA), and peripheral body temperature. We utilized short recordings from 9:00 to 9:15 pm and compared mean values between patients with and without impending seizures. In addition, we collected clinical data: age, sex, age at first seizure, generalized slowing, focal slowing, and spikes on EEG, magnetic resonance imaging findings, and antiseizure medication reduction. We used conventional machine learning techniques with cross-validation to classify patients with and without impending seizures.
Results |
We included 139 patients: 78 had no seizures and 61 had at least one seizure after 9 pm during the concurrent video-EEG and E4 recordings. HR (P < 0.01) and EDA (P < 0.01) were lower and HRV (P = 0.02) was higher for patients with than for patients without impending seizures. The average accuracy of group classification was 66%, and the mean area under the receiver operating characteristics was 0.72.
Conclusions |
Short-term wearable recordings in combination with clinical data have great potential as an easy-to-use seizure likelihood assessment tool.
Le texte complet de cet article est disponible en PDF.Keywords : Seizure forecasting, Supervised learning, Pediatric patients, Wearables, Clinical data
Plan
| Author contributions: S.V., W.J.B., and T.L. contributed to the study conception and design. Wearable data selection and clinical data collection were performed by S.C. and M.J. S.V. and W.J.B. designed the data analysis and S.V. analyzed the data. All authors interpreted the results, drafted parts of the work, approved the final version of the manuscript, and agreed to be accountable for all aspects of the work. |
|
| Funding: The study was supported by the German Research Foundation (DFG) under the grant VI 1088/1-1 and the Epilepsy Research Fund. W.B. was supported by a grant from the Koret Foundation. |
|
| Data availability statement: All statistical analyses and results are included in the article. Codes are not shared, but the packages used are listed and a link is provided in the article. The original data are available upon reasonable request and when compatible with the BCH IRB requirements. |
Vol 148
P. 118-127 - novembre 2023 Retour au numéro
Article précédent
- Perinatal Stroke in a Chinese Neonatal Center: Clinical Characteristics, Long-Term Outcomes, and Prognostic Factors
- Qianqian Xia, Fa Guo, Xinlin Hou, Zezhong Tang, Lili Liu
- Neurocognitive Profiles in Adolescents With Postural Tachycardia Syndrome and Perceived Brain Fog: A Preliminary Report
- Sarah A. Koch, Imad T. Jarjour, Karen D. Evankovich
Bienvenue sur EM-consulte, la référence des professionnels de santé.
L’accès au texte intégral de cet article nécessite un abonnement.
Déjà abonné à cette revue ?