Graduation Semester and Year
Summer 2025
Language
English
Document Type
Dissertation
Degree Name
Doctor of Philosophy in Electrical Engineering
Department
Electrical Engineering
First Advisor
Dr. Venkat Devarajan
Second Advisor
Dr. Ganesh Sankaranarayanan
Abstract
The rapid growth of surgical video analysis presents a need for efficient deep learning models for surgical training, while reducing the need for excessive image and video image storage. Traditional training approaches typically rely on uniformly compressed video data, instead of selectively preserving the most surgically relevant regions. This dissertation investigates the impact of training deep neural networks (DNNs), both convolutional and transformer-based architectures, on saliency-guided, differentially compressed surgical video sequences. The study systematically evaluates how such compression influences prediction accuracy, computational efficiency and storage requirements. Experimental results demonstrate that models trained on ROI-focused compressed data combined with motion vectors achieve comparable, and in some cases, improved, predictive performance for surgical tool detection relative to those trained on uniformly compressed inputs. Notably, these models may also benefit from reduced training time and lower storage demands, indicating the feasibility of integrating such methods into resource-constrained surgical workflows. This work highlights the practical potential of saliency-aware compression combined with motion vectors for enhanced prediction of surgical tools.
Keywords
Deep Learning in Surgery, Vision Transformers (ViT/DeiT), Frame-level Surgical Tool Detection, Multiclass Multilabel Classification, Compressed Domain Learning, Motion Aware Models, Transfer Learning, Motion Vectors, Multi-head Attention
Disciplines
Other Electrical and Computer Engineering | Signal Processing
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Recommended Citation
Sengupta, Aileen, "STUDY OF DEEP NEURAL NETWORK TRAINED WITH SALIENT, COMPRESSED MEDICAL VIDEO DATA FOR ENHANCED PREDICATION" (2025). Electrical Engineering Dissertations. 411.
https://mavmatrix.uta.edu/electricaleng_dissertations/411