AI-Enabled Digital Pathology for Precision Oncology

Artificial Intelligence has become the pivotal research and innovation vehicle in health care.

An increasing number of algorithms find their way into the clinical practice providing powerful solutions and assisting medical doctors in their everyday practice. The ambition of this grant is to further explore the field of deep-learning-based approaches and to propose novel, unbiased, and highly generalizable artificial intelligence algorithms for cancer treatment and response to immunotherapy. The main methodological contributions will focus on i) different learning schemes for training on gigapixel histopathological slides taking into account the spatial distribution of their tissue towards association to known biomarkers, ii) transformer-based architectures with different attention schemes for the fusion of histopathology and genetic/ clinical information towards biomarker discovery from important regions for the model's reasoning and iii) bias identification and domain adaptation methods based on the image to image translation and adversarial attacks for addressing domain shifts and possibly biological and clinical biases. Such an ambitious effort is highly interdisciplinary, as it lies at the interface between cancer care and artificial intelligence towards precision medicine and biomarker discovery.