Predicting Amyotrophic Lateral Sclerosis Progression: an EMG-based Survival Analysis

Irene Rechichi, Gianluca Amprimo, Alessandro Cicolin, Gabriella Olmo

Abstract: Amyotrophic Lateral Sclerosis (ALS) is a progressive neurodegenerative disease, ultimately leading to muscle inefficiency and death. A vast majority of people with ALS also suffer from sleep disorders. Previous studies highlighted the presence of REM Sleep Without Atonia (RSWA) in an ALS cohort, and suggested its strong correlation with the disease severity. This study investigates the ability of electromyography (EMG) parameters recorded during Rapid-eye Movement (REM) sleep to predict disease progress and outcome rapidity in ALS. Survival models trained on a cohort of 45 ALS patients undergoing a longitudinal study, revealed a promising predictive power for the proposed EMG-derived metrics (c-index ≥ 0.65) and encouraging goodness of fit (through c-index and χ2). These results suggest the possibility of employing the trained model in follow-up procedures, based on non-invasive, lightweight EMG metrics, which would significantly ease disease monitoring and help personalized symptomatic care.