{"id":31402,"date":"2026-06-16T13:06:45","date_gmt":"2026-06-16T11:06:45","guid":{"rendered":"https:\/\/sano.science\/?post_type=seminars&p=31402"},"modified":"2026-07-14T15:12:54","modified_gmt":"2026-07-14T13:12:54","slug":"195-learning-beyond-the-training-distribution-geometry-robust-augmentation-and-conditional-learning-in-medical-imaging","status":"publish","type":"seminars","link":"https:\/\/sano.science\/seminars\/195-learning-beyond-the-training-distribution-geometry-robust-augmentation-and-conditional-learning-in-medical-imaging\/","title":{"rendered":"195. Learning Beyond the Training Distribution: Geometry, Robust Augmentation, and Conditional Learning in Medical Imaging"},"content":{"rendered":"\n

Abstract:<\/h2>\n\n\n\n
<\/div>\n\n\n\n

Deep learning models often rely on dataset-specific patterns, limiting their ability to generalize to new domains and acquisition settings. This seminar presents a line of research conducted during my PhD at the Sano Centre for Computational Medicine that investigates how incorporating additional structure into the learning process can improve robustness and generalization. I will discuss three complementary perspectives: geometry-aware regularization, which constrains models using anatomical and geometric priors; robust data augmentation, which enriches information embedded in training data with prior knowledge; and conditional learning, where auxiliary shape representations are used to guide model predictions. Through examples from recent work, including geometry-based learning objectives, augmentation-driven robustness studies, and decoder conditioning with shape-derived embeddings, I will show how these approaches encourage models to learn representations that are less dependent on dataset-specific characteristics and more transferable to unseen data. Together, these studies illustrate how explicit inductive biases can improve the reliability of deep learning systems in medical imaging.<\/p>\n\n\n\n

<\/div>\n\n\n\n

About the author<\/h2>\n\n\n\n

Tomasz Szczepa\u0144ski<\/a> is a PhD candidate at the Sano Centre for Computational Medicine, where he is a member of the Medical Imaging and Robotics<\/a> group. His doctoral research, supervised by Prof. Tomasz Trzci\u0144ski (WUT) and co-supervised by Dr. Arkadiusz Sitek (Harvard Medical School), focuses on medical imaging, multimodal data integration, and geometric approaches to 3D image segmentation and generation.
His work has been presented at MICCAI (2023\u20132025) and ICCS (2022), and published in IEEE Transactions on Medical Imaging and Medical Image Analysis (2025). In 2025 he received an Outstanding Reviewer Award (Honorable Mention) at MICCAI, the only reviewer affiliated with a Polish institution to receive this distinction that year, following his service as a MICCAI reviewer since 2024.
He holds an MSc in Computer Science from WUT (2022, summa cum laude), for which his thesis on chest X-ray bias analysis in COVID-19 patients received a distinction in the National Master 4 Science competition, and a BEng in Photonics Engineering and Mechatronics, also from WUT (2018).<\/p>\n\n\n\n

<\/div>\n\n\n\n
\n