MedPath

Development of an Epileptic Seizure Detection Algorithm by Continuous Analysis of the Electrocardiogram

Conditions
Seizures
Electrocardiography
Epilepsy
Registration Number
NCT03369522
Lead Sponsor
University Hospital, Lille
Brief Summary

The aim of our project is to develop an epileptic seizure detection algorithm based on the the continuous analysis of the Electrocardiogram

Detailed Description

In a first step, the team will build a multi-parametric algorithm on existing records issued from the Epilepsy Monitoring Units. Each selected record will be analyzed by EMU medical team in order to detect seizure on the EEG. The algorithm will adapt in order to obtain the best sensitivity and specificity regarding seizure detection.

in a second step,algorithm will validate on a cohort of other recordings in patients with epilepsy

Recruitment & Eligibility

Status
UNKNOWN
Sex
All
Target Recruitment
200
Inclusion Criteria
  • patients with epilepsy
  • with a long-term video-EEG recording
  • 1 (or more) seizure (s) recorded
  • onset seizure clearly defined on the basis of the EEG
Exclusion Criteria
  • bad quality of the Electrocardiogram recording
  • pace-maker
  • non sinus rhythm

Study & Design

Study Type
OBSERVATIONAL
Study Design
Not specified
Primary Outcome Measures
NameTimeMethod
ROC curve for seizure detection24 months

ROC curve gives the relation between detection probability and false alarm probability of the algorithm that will be developped

Secondary Outcome Measures
NameTimeMethod
Time of detection of seizures24 months

Time of detection of seizures by the algorithm

ROC curve for Psychogenic seizure detection24 months

ROC curve for Psychogenic seizure detection

Trial Locations

Locations (4)

Ch de Valencienne

🇫🇷

Valenciennes, France

Chu Amiens Salouël

🇫🇷

Salouël, France

Ch Bethune

🇫🇷

Béthune, France

Hôpital Roger Salengro, CHU

🇫🇷

Lille, France

© Copyright 2025. All Rights Reserved by MedPath