An improved procedure for the extraction of temporal motion strength signals from video recordings of neonatal seizures

Vision-based human motion analysis in epilepsy - Methods and challenges

“…Considering marker-free methods, a series of publications from Karayiannis et al [2 1 ]- [28] report the application of image processing and pattern recognition techniques in differentiating between myoclonic seizures, focal clonic seizures and normal behavior in neonates. Analog video recordings from one camera were digitized at 30 fps using a resolution of 352 x 240 pixels.…”

“…These four features were: variance of time intervals between two adjacent spikes, energy ratio of the autocorrelation sequence, defined as the energy contained by the last 75% of the samples of the autocorrelation sequence to the energy contained by the first 25% of the samples, maximum spike duration and number of spikes/extrema per time unit [24]. The same authors also worked on a new MSS extraction procedure including nonlinear filtering, clustering and morphological filtering on the difference image between adjacent frames [28], as well as on new MAS extraction methods using adaptive block matching, various minimization approaches and a variety of block motion models. Further on, the authors tested and evaluated different techniques for segmenting the area occupied by the neonate's moving body parts in terms of optical flow application in MSS extraction including direct thresholding, clustering of pixel velocities and clustering of the motion parameters obtained by fitting an affine model to the pixel velocities [26].…”

“…Regarding the disadvantages of marker-based systems and considering the work of Karayiannis et al [2 1]- [28], as well as the evolution in marker-free VBHMA [4]- [7], it is obvious that a marker-free approach is feasible and highly desirable. We have, nevertheless, identified several challenges that need to be successfully addressed, which are also identified in the latest survey in vision-based human action recognition [7].…”

Joint Wavelet Video Denoising and Motion Activity Detection in Multimodal Human Activity Analysis: Application to Video-Assisted Bioacoustic/Psychophysiological Monitoring

“…The above settlement is quite common to many temporal-filtering-based video denoising algorithms [17,37], with various modifications encountered ccording to the involved motion detection/estimation parameters. The noise estimations are also refined following the outcome of (12) and the J N∼4 (w i , w j , n) components are extracted similarly to the J N∼1 and J N∼3 matrices (10), (11). Both J S∼3 (w i , w j , n) and J N∼4 (w i , w j , n) signals would be further utilized at the next iteration (processing at (n + 1) frame).…”

“…), (c) detection and isolation of movement artefacts that affect the integrity of the psychophysiological data, (d) validation and verification of various healthrelated symptoms/events, such as cough, apnoea episodes, restless leg syndrome, and so forth [1][2][3][4][5][6][7]. The majority of the video-assisted biomedical monitoring systems are engaged in polysomnography recordings during sleep studies [2][3][4][5][6][7], in various neurophysiology and kinesiology-related studies [8][9][10], for the extraction of temporal motion strength signals from video recordings of neonatal seizures [11]. Video monitoring and analysis allows physicians to evaluate the exact experimental condition under which the biomedical data were acquired [1].…”