Towards Efficient Models for Continual Learning in Medical Imaging
Date of Award
4-30-2024
Document Type
Thesis
Degree Name
Master of Science in Computer Vision
Department
Computer Vision
First Advisor
Prof. Fahad Khan
Second Advisor
Prof. Merouane Debbah
Abstract
Continual Learning (CL) in Deep Learning (DL) has emerged as a promising solution to address the challenges encountered in the practical applications of medical imaging systems. While DL has shown great potential in reshaping medical imaging, issues like catastrophic forgetting and domain shifts hinder its real-world adoption. In this regard, CL offers a solution by enabling models to sequentially acquire new knowledge without forgetting the previous. However, there exists a gap in the literature regarding a comprehensive review of CL within the context of medical imaging. Our work aims to fill this gap by providing an exhaustive review of CL methodologies, principles, and challenges in the medical domain. We delve into various CL approaches and their relevance in overcoming the gap between research and practical applications in medical imaging. Building upon the insights gained from the review, we propose Dynamic Model Merging, DynaMMo, a method that merges multiple networks at different stages of model training to achieve better computational efficiency. Specifically, we employ lightweight learnable modules for each task and combine them into a unified model to minimize computational overhead. DynaMMo achieves this without compromising performance, offering a cost effective solution for continual learning in medical applications. We evaluate DynaMMo on three publicly available datasets, demonstrating its effectiveness compared to existing approaches. DynaMMo offers around 10-fold reduction in GFLOPS with a small drop of 2.76 in average accuracy when compared to state-of-the-art dynamic-based approaches.
Recommended Citation
M. Qazi, "Towards Efficient Models for Continual Learning in Medical Imaging,", Apr 2024.
Comments
Thesis submitted to the Deanship of Graduate and Postdoctoral Studies
In partial fulfilment of the requirements for the M.Sc degree in Computer Vision
Advisors: Mohammad Yaqub, Karthik Nandakumar
with 2 years embargo period