Digitizing Movement for Differentiating Eeg-negative Epileptic Seizures from Non-epileptic Episodes in an Epilepsy Monitoring Unit (EMU)
Abstract number :
1.557
Submission category :
4. Clinical Epilepsy / 4B. Clinical Diagnosis
Year :
2024
Submission ID :
1510
Source :
www.aesnet.org
Presentation date :
12/7/2024 12:00:00 AM
Published date :
Authors :
Presenting Author: Charlotte Hooker, – Summa Health, Case Western Reserve University
Caroline Pylypyuk, BS – Summa Health
Leopoldo Zendejas-Zaragozal, PhD – Summa Health
Marvin Rossi, MD, PhD – Summa Health
Rationale: Electroencephalography (EEG)-negative epileptic seizures (ENES) & non-epileptic episodes (NEE) can present significant challenges in clinical diagnosis & management. This study explores the integration of motion capture wearable technology (MoCap; XSens) with video-EEG monitoring as a method for developing kinematic biomarkers between conditions, thus improving diagnostic accuracy in an EMU. Movement patterns were associated with a frontal lobe hypermotor ENES, a psychogenic non-epileptic seizure (PNES), & a subcortical movement disorder (SMD). Biomarkers included deviations in joint angles, acceleration, & velocity of movement, all relative to the established baseline kinematic profile. These biomarkers represented measurable indicators to facilitate the potential identification of the fragile neural network responsible for each condition.
Methods: Three subjects were included in the study. Respective histories included, subject 1: frequent paroxysmal spells following a remote closed brain injury, subject 2: post-traumatic stress disorder, & subject 3: sudden-onset episodes post coronary artery bypass surgery without anoxia. Duration of monitoring per subject = 2-4 hours. All subjects signed a consent prior to data collection. 17 sensors were applied to each participant in the EMU (shoulders, head, sternum, head, upper arms, forearms, & hands). Each sensor contained a 3D-gyroscope, accelerometer, & magnetometer to measure rotation rate, linear acceleration along each axis, & magnetic field strength relative to the Earth's magnetic field to correct for drift, ensuring accurate tracking of the body's position in space. Data acquisition was processed in real-time. This 3D model "digital skeleton" mirrored the participants' movements captured by video. These data were synchronized with EEG monitoring by aligning MoCap timestamps.
Results: Preliminary data identified episodes in all three subjects. Hypotheses for localization of choreiform-ballismus movements in the coronary bypass subject were consistent with a lesion in the basal ganglia. For the PNES subject, accelerometers detected habitual hand tremors consistent with PNES criteria, supporting the hypothesis that specific movements can be used as biomarkers to differentiate NES from epileptic seizures. Baseline data were collected, & the "digital skeleton" mirrored the participants' movements.
Conclusions: This study demonstrates the potential of integrating motion capture wearables with video-EEG monitoring in an EMU to enhance the differentiation between conditions. Detection of condition-specific kinematic biomarkers can distinguish between seizures, PNES, & choreiform-ballism SMD. Despite challenges related to cost, sensor calibration, & battery life, this technology shows promise for improving diagnostic accuracy & patient management.
Funding: n/a
Clinical Epilepsy