Video-EEG Data Compression

University Of Pittsburgh At Pittsburgh, Pittsburgh PA

Investigators

Linked publications & trials

Abstract

Prolonged simultaneous recording of both electroencephalogram (EEG) waveforms and video is often conducted during the evaluation of patients with seizures. Recently, digital video-EEG systems based on MPEG video compression standards have emerged. These systems can provide quick access to any video segment of interest, and support various display options on computer screens. However, they have sub-optimal data compression performance because the existing MPEG-based software packages, which mainly target applications such as films and digital TV, do not adapt well to the case of epilepsy video monitoring over extended periods of time. As a result, important applications such as data archiving and management, data access through the Internet, remote diagnosis, and home epilepsy monitoring have been hampered due to the excessive data size. We propose an investigation on video compression to be applied specifically to aid in epilepsy diagnosis. We will develop new algorithms for video object segmentation based on special characteristics of epilepsy video and the MPEG-4 video compression standard. Using these algorithms we will design a state-of-the-art high-resolution, low output rate epilepsy data acquisition system for both EEG and video to support rapid Internet data transmission and efficient data archiving. Finally, we will conduct a series of field-tests at remote hospital sites in rural regions to evaluate our system.