Seizure detection using wearable electrocardiogram connected to a smartphone: a phase 3 clinical validation study.

IF 10.8 1区医学 Q1 MEDICINE, RESEARCH & EXPERIMENTAL

EBioMedicine Pub Date : 2025-10-01 Epub Date: 2025-09-29 DOI:10.1016/j.ebiom.2025.105952

Jesper Jeppesen, Jakob Christensen, Oliver Ahrenfeldt Petersen, Sarah Fenger, Sidsel Armand Larsen, Stephan Wüstenhagen, Stefan Rahr Wagner, Peter Johansen, Sándor Beniczky

{"title":"Seizure detection using wearable electrocardiogram connected to a smartphone: a phase 3 clinical validation study.","authors":"Jesper Jeppesen, Jakob Christensen, Oliver Ahrenfeldt Petersen, Sarah Fenger, Sidsel Armand Larsen, Stephan Wüstenhagen, Stefan Rahr Wagner, Peter Johansen, Sándor Beniczky","doi":"10.1016/j.ebiom.2025.105952","DOIUrl":null,"url":null,"abstract":"Background: Automated seizure detection is needed for patient safety and for objective seizure quantification. Wearable seizure detection devices hold great potential to improve patient care. Our objectives were to assess the accuracy of a wearable ECG-device connected to a smartphone, in detecting epileptic seizures in patients with autonomic ictal changes, and evaluate its capability to automatically determine impairment of consciousness.Methods: We conducted a phase 3, prospective, blinded, multicentre, clinical validation study of real-time seizure detection using a predefined algorithm. We recruited consecutive patients admitted to Epilepsy Monitoring Units. Eligible patients experienced seizures with autonomic ictal manifestations, defined as ictal heart rate change exceeding 50 beats per minute, inferred from the first recorded seizure. Patients wore an ECG-device connected to a smartphone. The algorithm, based on heart rate variability, used a personalised detection threshold determined from the first 24 h of recording. During daytime, seizure detection triggered automated behavioural-testing on the smartphone to confirm detection and assess consciousness.Findings: Of 101 enrolled patients, 36 experienced seizures, with 42 seizures recorded from 17 eligible patients. Overall sensitivity across all 42 seizures was 90·5% (95% CI: 77·4-97·3%), median sensitivity per patient was 100% (95% CI: 100-100%). All bilateral tonic-clonic seizures were detected, while sensitivity for other focal seizures was 82·6% (95% CI: 61·2-95·1%), median per patient: 100% (95% CI: 60-100%). Mean false alarm rate was 2·5/day (median per patient: 1·1/day, 95% CI: 0-2·8/day, zero during the night). Device deficiency time was 1·8% and signal loss was 4·5% (median per patient: 0·3% and 0·5% respectively). Use of the behavioural-testing application successfully cancelled all false alarms and accurately identified impairment of consciousness.Interpretation: The wearable ECG device connected to a smartphone accurately detected focal and generalised seizures, and assessed impairment of consciousness.Funding: Independent Research Fund Denmark (grant number 0134-00400B).","PeriodicalId":11494,"journal":{"name":"EBioMedicine","volume":"120 ","pages":"105952"},"PeriodicalIF":10.8000,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12516532/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"EBioMedicine","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1016/j.ebiom.2025.105952","RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2025/9/29 0:00:00","PubModel":"Epub","JCR":"Q1","JCRName":"MEDICINE, RESEARCH & EXPERIMENTAL","Score":null,"Total":0}

引用次数: 0

Abstract

Background: Automated seizure detection is needed for patient safety and for objective seizure quantification. Wearable seizure detection devices hold great potential to improve patient care. Our objectives were to assess the accuracy of a wearable ECG-device connected to a smartphone, in detecting epileptic seizures in patients with autonomic ictal changes, and evaluate its capability to automatically determine impairment of consciousness.

Methods: We conducted a phase 3, prospective, blinded, multicentre, clinical validation study of real-time seizure detection using a predefined algorithm. We recruited consecutive patients admitted to Epilepsy Monitoring Units. Eligible patients experienced seizures with autonomic ictal manifestations, defined as ictal heart rate change exceeding 50 beats per minute, inferred from the first recorded seizure. Patients wore an ECG-device connected to a smartphone. The algorithm, based on heart rate variability, used a personalised detection threshold determined from the first 24 h of recording. During daytime, seizure detection triggered automated behavioural-testing on the smartphone to confirm detection and assess consciousness.

Findings: Of 101 enrolled patients, 36 experienced seizures, with 42 seizures recorded from 17 eligible patients. Overall sensitivity across all 42 seizures was 90·5% (95% CI: 77·4-97·3%), median sensitivity per patient was 100% (95% CI: 100-100%). All bilateral tonic-clonic seizures were detected, while sensitivity for other focal seizures was 82·6% (95% CI: 61·2-95·1%), median per patient: 100% (95% CI: 60-100%). Mean false alarm rate was 2·5/day (median per patient: 1·1/day, 95% CI: 0-2·8/day, zero during the night). Device deficiency time was 1·8% and signal loss was 4·5% (median per patient: 0·3% and 0·5% respectively). Use of the behavioural-testing application successfully cancelled all false alarms and accurately identified impairment of consciousness.

Interpretation: The wearable ECG device connected to a smartphone accurately detected focal and generalised seizures, and assessed impairment of consciousness.

Funding: Independent Research Fund Denmark (grant number 0134-00400B).

Abstract Image

查看原文本刊更多论文

使用连接到智能手机的可穿戴心电图检测癫痫发作：一项3期临床验证研究。

背景：为了患者安全和客观的癫痫定量，需要自动检测癫痫发作。可穿戴式癫痫检测设备在改善患者护理方面具有巨大潜力。我们的目的是评估与智能手机连接的可穿戴心电图设备在检测自主神经变化患者癫痫发作方面的准确性，并评估其自动判断意识损伤的能力。方法：我们进行了一项使用预定义算法实时检测癫痫发作的3期、前瞻性、盲法、多中心临床验证研究。我们招募了连续入住癫痫监护病房的患者。符合条件的患者有癫痫发作的自主神经表现，定义为从首次记录的癫痫发作推断的癫痫心率变化超过每分钟50次。患者戴着与智能手机相连的心电图设备。该算法基于心率变异性，使用从记录的前24小时确定的个性化检测阈值。在白天，癫痫检测触发智能手机上的自动行为测试，以确认检测并评估意识。结果：101例入组患者中，36例发生癫痫发作，17例符合条件的患者中记录了42例癫痫发作。42次癫痫发作的总敏感性为99.5% (95% CI: 77.4 - 97.3%)，每位患者的中位敏感性为100% （95% CI: 100-100%）。所有双侧强直-阵挛性发作均被检测到，而其他局灶性发作的敏感性为82.6% (95% CI: 62.1 - 95.1%)，每位患者中位数为100% （95% CI: 60-100%）。平均虚警率为2.5次/天（每位患者中位数：1.1次/天，95% CI: 0- 2.8次/天，夜间为零）。器械缺陷时间为1.8%，信号丢失为4.5%（每位患者中位数分别为0.3%和0.5%）。使用行为测试应用程序成功地取消了所有假警报，并准确地识别了意识损伤。解释：与智能手机连接的可穿戴ECG设备可准确检测局灶性和全身性癫痫发作，并评估意识损害。资助：丹麦独立研究基金（资助号0134-00400B）。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

EBioMedicine Biochemistry, Genetics and Molecular Biology-General Biochemistry,Genetics and Molecular Biology

CiteScore

17.70

自引率

0.90%

发文量

579

审稿时长

5 weeks

期刊介绍： eBioMedicine is a comprehensive biomedical research journal that covers a wide range of studies that are relevant to human health. Our focus is on original research that explores the fundamental factors influencing human health and disease, including the discovery of new therapeutic targets and treatments, the identification of biomarkers and diagnostic tools, and the investigation and modification of disease pathways and mechanisms. We welcome studies from any biomedical discipline that contribute to our understanding of disease and aim to improve human health.