Vasiliki Sideri-Lampretsa

I'm a PhD student and researcher in Computer Science at the Technical University of Munich (TUM) and TUM Klinikum Rechts der Isar. I work at the intersection of medical imaging, neural fields, and representation learning. My research focuses on implicit neural representations and uni- and multi-modal medical image registration.

Prior to my PhD, I earned my Diploma in Electrical and Computer Engineering from the Aristotle University of Thessaloniki and my M.Sc. in Computer Science from the Technical University of Munich (TUM), where my thesis explored ultrasound simulations for improved image reconstruction.

Email  /  Scholar  /  Github  /  LinkedIn

Research (Selected Publications)

I'm interested in machine learning, neural compression and signal modeling. Some papers are highlighted.

We give the first mechanistic account of what transfers in cohort-trained INRs: the optimal freeze depth coincides with the layer of highest weight stable rank, and SAE decompositions reveal that SIREN learns localized, coordinate-tiling atoms while FFMLP learns image-spanning atoms that memorize cohort content.

We propose using sparse autoencoders to learn interpretable representations of medical images, enabling structured and human-understandable feature discovery in clinical imaging data.

We develop a holistic Lipschitz perspective for interpreting the learning process and the regularization of implicit neural representations.

We leverage diffusion models to generate deformation fields for building conditional atlases, enabling flexible and high-quality anatomical atlas construction conditioned on subject-specific attributes.

We introduce SINR, a method to parameterize the continuous deformable transformation represented by an INR using Free Form Deformations (FFD) enabling robust multi-modal registration while preventing spatial folding.

We leverage large-scale UK Biobank whole-body MRI data to construct population-specific atlases, capturing demographic and biological variation across diverse subgroups.

We investigate deep learning approaches for pulse-echo sound speed estimation in breast ultrasound, exploring the potential of neural networks to improve quantitative acoustic imaging.

Off-Grid: 1st Workshop on Continuous Representations and Grid-Free Methods in Medical Imaging (MICCAI 2026)

Paul Friedrich, Julian McGinnis, Vasiliki Sideri-Lampretsa, Yunjie Chen, Florentin Bieder, Suprosanna Shit, Jelmer M. Wolterink

MICCAI 2024 Tutorial on Implicit Neural Representations for Medical Imaging

Julian McGinnis, Nil Stolt-Ansó, Veronika Spieker, Suprosanna Shit, Maik Dannecker, Vasiliki Sideri-Lampretsa, Björn Menze, Daniel Rückert, Benedikt Wiestler

Based on Jon Barron's website.