Theodoros Pissas

I am a postdoctoral researcher at the University of Bern in Switzerland, working with Prof. Raphael Sznitman within the Artificial Intelligence in Medical Imaging group. My current research focuses on large-scale multimodal learning for generalizable medical image understanding and on semantic segmentation in natural and medical domains.

Prior to joining UniBern, I completed my PhD at University College London under the supervision of Prof. Christos Bergeles and Prof. Lyndon Da Cruz working on pixel-level semantic understanding of multimodal ophthalmic images and natural scenes.

A long time ago, I studied Electrical and Computer Engineering at the National Technical University of Athens where I worked on action recognition using video and depth sensors under the supervision of Prof. Petros Maragos

Email / CV / Google Scholar / Twitter / Github

Research

I'm generally interested in all aspects of recognition problems (segmentation, detection, classification) in both natural and medical domains. I am also passionate about developing self-supervised learning methods using large-scale multimodal medical data from visual and textual sources. My current research also aims to develop foundational multimodal models for ophthalmology that can be easily adapted to perform various tasks such as biomarker detection, segmentation or disease progression estimation.

	Stochastic Segmentation with Conditional Categorical Diffusion Models Lukas Zbinden, Lars Doorenbos, Theodoros Pissas, Adrian Thomas Huber, Raphael Sznitman, Pablo Márquez-Neila, ICCV, 2023 arXiv / code Conditional Categorical Diffusion models are effective in learning multi-rater segmentation distributions and standard semantic segmentation.
	Multi-scale and Cross-scale Contrastive Learning for Semantic Segmentation Theodoros Pissas, Claudio Ravasio, Lyndon Da Cruz, Christos Bergeles, ECCV , 2022 arxiv / code / video Multi-scale and Cross-scale contrastive learning improves semantic segmentation for many standard architectures and datasets.
	Effective semantic segmentation in Cataract Surgery: What matters most? Theodoros Pissas, Claudio Ravasio, Lyndon Da Cruz, Christos Bergeles, MICCAI, 2021 (Ranked 1st at MICCAI Endoscopic Vision (EndoVis) Challenge 2020 ) arxiv / code / demo / video Addressing class imbalance for semantic segmentation in Cataract Surgery videos is essential for effective surgical scene understanding
	Unpaired Intra-operative OCT (iOCT) Video Super-Resolution with Contrastive Learning Charalampos Komninos, Theodoros Pissas, Blanca Flores, Edward Bloch, Tom Vercauteren, Sébastien Ourselin, Lyndon Da Cruz, Christos Bergeles, Biomedical Optics Express, 2023 paper / demo Video Super Resolution approach to improve the quality of Intra-operative OCT during vitreo-retinal surgery.
	Surgical biomicroscopy-guided intra-operative optical coherence tomography (iOCT) image super-resolution Charalampos Komninos, Theodoros Pissas, Lina Mekki, Edward Bloch, Blanca Flores, Tom Vercauteren, Sébastien Ourselin Lyndon Da Cruz, Christos Bergeles, International Journal of Computer Assisted Radiology and Surgery (IJCARS) , 2022 paper
	Deep iterative vessel segmentation in OCT angiography Theodoros Pissas, Edward Bloch, Blanca Flores, Odysseas Georgiadis, Sepehr Jalali, Claudio Ravasio, Danail Stoyanov, Lyndon Da Cruz, Christos Bergeles, Biomedical Optics Express, 2020 (Editor's pick) paper / code Blood vessel segmentation in Optical Coherence Tomography Angiography of the human retina.
	Learned optical flow for intra-operative tracking of the retinal fundus Claudio Ravasio, Theodoros Pissas, Edward Bloch, Blanca Flores, Odysseas Georgiadis, Sepehr Jalali, Danail Stoyanov, Jorge M Cardoso, Lyndon Da Cruz, Christos Bergeles, International Journal of Computer Assisted Radiology and Surgery (IJCARS) and IPCAI 2020 paper \ video Learning optical flow for intraoperative tracking of the retinal fundus in real-world retinal surgery using synthetic data.
	Pixel-level semantic understanding of ophthalmic images and beyond Theodoros Pissas, University College London , 2022 thesis PhD thesis

I stole this template from Jon Barron.