Deep Learning in Cancer Research: Predicting Survival, Therapy Response, and Understanding Tumour Microenvironment – Nitigya Sambyal

Deep learning has emerged as a transformative tool in cancer research, offering advanced techniques to interpret complex biomedical data and provide novel solutions to clinical decision making. With the ability to learn hierarchical features and detect subtle patterns in data, deep learning has outperformed traditional methods in multiple domains, enabling researchers and clinicians to make more accurate decisions. This seminar presents an overview of the deep learning techniques specifically in three areas of cancer research namely, Survival prediction, Therapy response prediction, and Tumour microenvironment (TME) characterization. We begin by exploring how deep learning models, ranging from classical architectures like convolutional neural networks (CNNs) to more advanced approaches like graph neural networks (GNNs) and transformers are utilized for survival prediction. We also examine the growing use of deep learning in therapy response prediction, particularly in forecasting how individual patients respond to immunotherapies, chemotherapies etc. Finally, we delve into recent advancements in TME characterization, where deep learning is used to decode the spatial & cellular complexity of tumours, classify immune infiltration patterns, identify stromal subtypes, and cell-cell interactions. The seminar concludes by identifying emerging trends and future directions in cancer research. By highlighting key studies and challenges, this seminar aims to shed light on how AI is shaping the future of personalized cancer care.

About Nitigya Sambyal