Region of Interest-Based Hevc Video Transmission for Telemedicine in Healthcare

Abstract

Telemedicine plays crucial role in providing care to distant patients, necessitating efficient transmission of Region of Interest (RoI) data without interruptions. To ensure transmitting video over a limited bandwidth to enable uninterrupted data sharing. RoI-based solution that modified the HEVC to encode high-quality RoI in dynamic positions. Furthermore, the background of the image is blurred to reduce size and highlight the RoI. The flow of the underlying layers is improved for better RoI extraction. Our approach, facilitating transmission over low bandwidth. The default HEVC consumes 2929.3 kbps for patient and 8140.6 kbps for doctor. In contrast, the proposed solution reduces the bitrate to 735.2 kbps and 2757.8 kbps, respectively. In case of RoI with a QP value of 20, the default HEVC consumes a PSNR of 49.61 db at a bit rate of 5939.27 kbps whereas the proposed solution achieved 49.77 db at a reduced bit rate of 2552.19 kbps.

Keywords: Healthcare, low bandwidth, medical multimedia, deep learning