Advancing Deep Learning in Medical Imaging Through Generative Modeling and Representation Learning
MCML Authors
Tobias Weber
* Former Member
Abstract
Tobias Weber
* Former Member
Abstract
In recent years, deep learning (DL) has proven to be a disruptive enabler in many domains, including the realm of medical imaging. The application of neural networks and other learnable algorithms has substantially impacted the medical field, promising to improve diagnostic accuracy, enhance patient outcomes, and streamline clinical workflows. The advent of large-scale datasets and advancements in computational power have facilitated the development of sophisticated DL models capable of analyzing and interpreting complex medical images. The scope of this thesis concentrates on a subset of the full DL spectrum, specifically the uprising areas of generative modeling and representation learning, which are closely interleaved with each other. The proposed contributions aim to push the boundaries of established medical image DL methods, venturing into more experimental research areas. (Shortened)
BibTeXKey: Web25