Enhancing Extended Reality Assisted Surgery through a Field-of-View Video Delivery Optimization

Emerging Extended Reality (XR) applications bring new opportunities for digital healthcare systems, i.e., eHealth. XR-assisted surgery is one of the most outstanding examples of future technology that has a high social impact on the healthcare and medical educational system. The current work presents the intelligent design for remote XR-assisted surgery. The study presents the Field-of-View (FoV)-based viewport model empowered with behavioral data. It applies the viewport prediction model based on the behavioral data by applying Artificial Neural Network (ANN) and Long Short-Term Memory (LSTM). In the final analysis, LSTM showed lower errors and a higher coefficient of determination, but ANN performed much faster. Finally, the study defines the dynamic system’s states for adaptive and fast video delivery concerning Quality of Experience (QoE). The presented approach aims to mitigate the delay to ensure smooth playback and display high-quality images. • Explores advanced techniques for mitigating video traffic. • Presents a video delivery model for an advanced dynamic tiled-based video transmission. • Outlines an open-source datasets applicable to XR-related research. • Presents the results on behavioral prediction by applying ANN and LSTM. • Emphasizes the importance of behavioral study and the state of system dynamics in XR.