Extraction of motion strength and motor activity signals from video recordings of neonatal seizures

This paper presents two methods developed to extract quantitative informati on from video recordings of neonatal seizures in the form of temporal motio n strength and motor activity signals. Motion strength signals are extracte d by measuring the area of the body parts that move during the seizure and the relative speed of motion using a combination of spatiotemporal subband decomposition of video, nonlinear filtering, and segmentation. Motor activi ty signals are extracted by tracking selected anatomical sites during the s eizure using a modified version of a feature-tracking procedure developed f or video, known as the Kanade-Lucas-Tomasi (KLT) algorithm. The experiments indicate that the temporal signals produced by the proposed methods provid e the basis for differentiating myoclonic from focal clonic seizures and di stinguishing these types of neonatal seizures from normal infant behaviors.