Quality evaluation of medical ultrasound videos for e-health and telemedicine applications

Razaak, Manzoor (2015) Quality evaluation of medical ultrasound videos for e-health and telemedicine applications. (PhD thesis), Kingston University, .


The advancements in multimedia communication technologies have enabled an increased implementation of telemedicine and e-health application for healthcare services. In parallel, advanced imaging methods have facilitated increasing reliance on medical images and videos for patient diagnosis. The high data speeds achieved by current communication technologies enables reliable transmission of medical videos for diagnosis and education purposes in telemedicine applications. The necessary process of video compression, prior to transmission, and communication channel constraints may occasionally impact the quality of the medical video received after transmission. Thus, to verify the reliability of the received video, quality evaluation is necessary. However, the present approaches used for medical video quality evaluation have limitations in addressing the contextual requirements of medical videos. The research work presented in this thesis addresses quality evaluation of medical ultrasound videos for telemedicine and e-health applications. The studies presented in the thesis include a subjective quality assessment study of medical ultrasound videos compressed via the High Efficiency Video Coding (HEVC) standard and the validation of the performance of state-of-the-art video quality metrics using the subjective cores of medical experts. Further, the rate-distortion and rate-quality performance of HEVC is analysed for the compression of medical ultrasound videos. A video quality metric, Cardiac Ultrasound Quality Index (CUQI), for cardiac ultrasound videos is proposed that considers the motion and edge features of cardiac videos for quality evaluation. The proposed metric assessment closely agrees with the subjective assessment of medical experts. Finally, a content-aware packet scheduling approach for transmission of medical ultrasound videos over Long Term Evolution (LTE) wireless network is presented. The scheduling approach employs a utility function based on the temporal complexity of the medical ultrasound videos and results in improving the received video quality. The research outcomes presented in the thesis indicate that developing quality evaluation approaches according to the contextual requirements of the medical video modality could enable overcoming the limitations of standard quality evaluation approaches.

Actions (Repository Editors)

Item Control Page Item Control Page