Automated Epileptic Seizure Type Classification through Quantitative Movement Analysis

摘要

In this paper we present the evolution of a quantitative movement analysis methodology for epileptic seizures. With this improved method we analyzed 20 seizure video sequences, 10 classified as automotor and 10 as hypermotor, from 17 different patients. The results obtained show we could classify all (100%) of the hypermotor seizures solely based on a quantified movement parameter - called movement extent -extracted with our method. Other quantitative parameters were also studied. This striking result paves the way to the contribution of quantitative movement methods in automated epileptic seizure detection systems.