Deep anomaly detection of seizures with paired stereoelectroencephalography and video recordings

Abstract Real-time seizure detection is a resource intensive process as it requires continuous monitoring of patients on stereoelectroencephalography. This study improves real-time seizure detection in drug resistant epilepsy (DRE) patients by developing patient-specific deep learning models that ut...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Michael L. Martini, Aly A. Valliani, Claire Sun, Anthony B. Costa, Shan Zhao, Fedor Panov, Saadi Ghatan, Kanaka Rajan, Eric Karl Oermann
Formato: article
Lenguaje:EN
Publicado: Nature Portfolio 2021
Materias:
R
Q
Acceso en línea:https://doaj.org/article/d773d42d7cf9413393bd341e9a21fbe5
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:d773d42d7cf9413393bd341e9a21fbe5
record_format dspace
spelling oai:doaj.org-article:d773d42d7cf9413393bd341e9a21fbe52021-12-02T14:15:53ZDeep anomaly detection of seizures with paired stereoelectroencephalography and video recordings10.1038/s41598-021-86891-y2045-2322https://doaj.org/article/d773d42d7cf9413393bd341e9a21fbe52021-04-01T00:00:00Zhttps://doi.org/10.1038/s41598-021-86891-yhttps://doaj.org/toc/2045-2322Abstract Real-time seizure detection is a resource intensive process as it requires continuous monitoring of patients on stereoelectroencephalography. This study improves real-time seizure detection in drug resistant epilepsy (DRE) patients by developing patient-specific deep learning models that utilize a novel self-supervised dynamic thresholding approach. Deep neural networks were constructed on over 2000 h of high-resolution, multichannel SEEG and video recordings from 14 DRE patients. Consensus labels from a panel of epileptologists were used to evaluate model efficacy. Self-supervised dynamic thresholding exhibited improvements in positive predictive value (PPV; difference: 39.0%; 95% CI 4.5–73.5%; Wilcoxon–Mann–Whitney test; N = 14; p = 0.03) with similar sensitivity (difference: 14.3%; 95% CI − 21.7 to 50.3%; Wilcoxon–Mann–Whitney test; N = 14; p = 0.42) compared to static thresholds. In some models, training on as little as 10 min of SEEG data yielded robust detection. Cross-testing experiments reduced PPV (difference: 56.5%; 95% CI 25.8–87.3%; Wilcoxon–Mann–Whitney test; N = 14; p = 0.002), while multimodal detection significantly improved sensitivity (difference: 25.0%; 95% CI 0.2–49.9%; Wilcoxon–Mann–Whitney test; N = 14; p < 0.05). Self-supervised dynamic thresholding improved the efficacy of real-time seizure predictions. Multimodal models demonstrated potential to improve detection. These findings are promising for future deployment in epilepsy monitoring units to enable real-time seizure detection without annotated data and only minimal training time in individual patients.Michael L. MartiniAly A. VallianiClaire SunAnthony B. CostaShan ZhaoFedor PanovSaadi GhatanKanaka RajanEric Karl OermannNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-11 (2021)
institution DOAJ
collection DOAJ
language EN
topic Medicine
R
Science
Q
spellingShingle Medicine
R
Science
Q
Michael L. Martini
Aly A. Valliani
Claire Sun
Anthony B. Costa
Shan Zhao
Fedor Panov
Saadi Ghatan
Kanaka Rajan
Eric Karl Oermann
Deep anomaly detection of seizures with paired stereoelectroencephalography and video recordings
description Abstract Real-time seizure detection is a resource intensive process as it requires continuous monitoring of patients on stereoelectroencephalography. This study improves real-time seizure detection in drug resistant epilepsy (DRE) patients by developing patient-specific deep learning models that utilize a novel self-supervised dynamic thresholding approach. Deep neural networks were constructed on over 2000 h of high-resolution, multichannel SEEG and video recordings from 14 DRE patients. Consensus labels from a panel of epileptologists were used to evaluate model efficacy. Self-supervised dynamic thresholding exhibited improvements in positive predictive value (PPV; difference: 39.0%; 95% CI 4.5–73.5%; Wilcoxon–Mann–Whitney test; N = 14; p = 0.03) with similar sensitivity (difference: 14.3%; 95% CI − 21.7 to 50.3%; Wilcoxon–Mann–Whitney test; N = 14; p = 0.42) compared to static thresholds. In some models, training on as little as 10 min of SEEG data yielded robust detection. Cross-testing experiments reduced PPV (difference: 56.5%; 95% CI 25.8–87.3%; Wilcoxon–Mann–Whitney test; N = 14; p = 0.002), while multimodal detection significantly improved sensitivity (difference: 25.0%; 95% CI 0.2–49.9%; Wilcoxon–Mann–Whitney test; N = 14; p < 0.05). Self-supervised dynamic thresholding improved the efficacy of real-time seizure predictions. Multimodal models demonstrated potential to improve detection. These findings are promising for future deployment in epilepsy monitoring units to enable real-time seizure detection without annotated data and only minimal training time in individual patients.
format article
author Michael L. Martini
Aly A. Valliani
Claire Sun
Anthony B. Costa
Shan Zhao
Fedor Panov
Saadi Ghatan
Kanaka Rajan
Eric Karl Oermann
author_facet Michael L. Martini
Aly A. Valliani
Claire Sun
Anthony B. Costa
Shan Zhao
Fedor Panov
Saadi Ghatan
Kanaka Rajan
Eric Karl Oermann
author_sort Michael L. Martini
title Deep anomaly detection of seizures with paired stereoelectroencephalography and video recordings
title_short Deep anomaly detection of seizures with paired stereoelectroencephalography and video recordings
title_full Deep anomaly detection of seizures with paired stereoelectroencephalography and video recordings
title_fullStr Deep anomaly detection of seizures with paired stereoelectroencephalography and video recordings
title_full_unstemmed Deep anomaly detection of seizures with paired stereoelectroencephalography and video recordings
title_sort deep anomaly detection of seizures with paired stereoelectroencephalography and video recordings
publisher Nature Portfolio
publishDate 2021
url https://doaj.org/article/d773d42d7cf9413393bd341e9a21fbe5
work_keys_str_mv AT michaellmartini deepanomalydetectionofseizureswithpairedstereoelectroencephalographyandvideorecordings
AT alyavalliani deepanomalydetectionofseizureswithpairedstereoelectroencephalographyandvideorecordings
AT clairesun deepanomalydetectionofseizureswithpairedstereoelectroencephalographyandvideorecordings
AT anthonybcosta deepanomalydetectionofseizureswithpairedstereoelectroencephalographyandvideorecordings
AT shanzhao deepanomalydetectionofseizureswithpairedstereoelectroencephalographyandvideorecordings
AT fedorpanov deepanomalydetectionofseizureswithpairedstereoelectroencephalographyandvideorecordings
AT saadighatan deepanomalydetectionofseizureswithpairedstereoelectroencephalographyandvideorecordings
AT kanakarajan deepanomalydetectionofseizureswithpairedstereoelectroencephalographyandvideorecordings
AT erickarloermann deepanomalydetectionofseizureswithpairedstereoelectroencephalographyandvideorecordings
_version_ 1718391752095694848