Postdoctoral Researcher · NKI & UvA · Amsterdam
AI for Oncology · Deep Learning · Computer Vision
AI for Oncology Lab · Netherlands Cancer Institute · University of Amsterdam
I am a postdoctoral researcher at the Netherlands Cancer Institute, supervised by Jonas Teuwen, working on deep learning and computer vision for AI in oncology. My main focus is medical image reconstruction; I also work on foundation models and collaborate on AI projects involving radiology, surgery, pathology, biology, and proteomics.
I lead the open-source DIRECT reconstruction toolkit and contribute to foundation model research through AIFOFOMO in the Foundation Models for Oncology Lab. I completed my PhD at the NKI and UvA on deep learning for MRI reconstruction, adaptive sampling, and motion estimation, supervised by Jonas Teuwen, Jan Jakob Sonke, and Clara I. Sánchez.
AI researcher in deep learning and computer vision for oncology — mainly reconstruction and foundation models, with collaborative work on AI for medical imaging and molecular data.
Postdoctoral researcher in the AI for Oncology Lab, supervised by Jonas Teuwen, working on reconstruction, foundation models (AIFOFOMO), and collaborative AI projects across imaging and molecular domains at NKI.
What We Do Sample, We Must Learn to Reconstruct: From Missing k-Space Data to Meaningful Images — Deep Learning in MRI Reconstruction and Beyond
Supervised by Jonas Teuwen, Jan Jakob Sonke, and Clara I. Sánchez. My thesis combined deep learning with inverse problems, image reconstruction, adaptive sampling, and motion estimation for accelerated MRI — with publications at CVPR, MICCAI, Medical Image Analysis, and Magnetic Resonance Imaging.
Recent updates from conferences, milestones, and ongoing research. Browse the complete archive on the news page.
Our new preprint MIRAGE: Multi-scale Lesion-Informed Representation with Auxiliary Guidance for MRI Contrast Enhancement is now on arXiv (arXiv:2607.19137), with Andrea Borghesi, Xin Wang, and Jonas Teuwen.
MIRAGE is a residual 2D U-Net with lesion-aware supervision for inferring contrast enhancement from pre-contrast breast MRI, improving lesion localization on the multi-centre MAMA-SYNTH dataset.
I am attending the 9th International Conference on Medical Imaging with Deep Learning (MIDL 2026) in Taipei, Taiwan, 8–10 July at the Chientan Youth Activity Center — the first MIDL edition held in Asia.
I am presenting two works as posters: End-to-End Co-Optimization of Adaptive k-Space Sampling and Reconstruction for Dynamic MRI and Conditional Learned Reconstruction for Medical Imaging (with Nikita Moriakov, Jan-Jakob Sonke, and Jonas Teuwen).
Team fomofo from the Foundation Models for Oncology Lab placed first on Task 1 (All-data Track) of the CVPR 2026 Challenge on Foundation Models for General CT Image Diagnosis, evaluating frozen pretrained 3D CT representations with linear probing across chest and abdomen.
Our submission using TAP-CT achieved the top balanced accuracy (0.654) and AUROC (0.714) on the public leaderboard — part of the foundation model work I contribute to through AIFOFOMO at NKI.
Lead developer · NKI-AI
Deep Image REConstruction Toolkit
Open-source PyTorch pipeline for deep learning-based medical image reconstruction — MRI reconstruction, denoising, and dealiasing. Ships RecurrentVarNet, vSHARP, RIM, LPD, and VarNet with pretrained models in the Model Zoo and configs for Calgary-Campinas, FastMRI, and CMRxRecon.
Winner · Multi-Coil MRI Challenge 2022 · Runner-up · CMRxRecon 2023 · 2nd & 3rd · CMRxRecon 2024
Peer-reviewed and preprint work on reconstruction, computer vision, and deep learning for medical imaging.
PhD thesis at the University of Amsterdam and Netherlands Cancer Institute on deep learning for accelerated MRI reconstruction — from missing k-space data and adaptive sampling to self-supervised learning, cross-domain generalization, and clinical validation. First doctoral graduate from the AI for Oncology Lab.
E2E-ADS-Recon jointly learns frame-specific adaptive k-space sampling and dynamic MRI reconstruction end-to-end — replacing fixed subsampling patterns with learned, temporally aware sampling that improves reconstruction quality across acceleration factors.
Retrospective, paired, multi-reader multi-case study at UMCG showing that AI-reconstructed prostate MRI at reduced acquisition times maintains PI-RADS-based prostate cancer detection comparable to conventional scans — a step toward faster prostate MRI without compromising diagnostic performance.
Netherlands Cancer Institute · AI for Oncology Lab · Amsterdam
Supervised by Jonas Teuwen. Reconstruction and foundation models (AIFOFOMO), plus collaborative AI work with radiology, pathology, surgery, biology, and proteomics colleagues.
Netherlands Cancer Institute & University of Amsterdam
Supervised by Jonas Teuwen, Jan Jakob Sonke, and Clara I. Sánchez. Deep learning for accelerated MRI acquisition, reconstruction, adaptive sampling, and real-time tumor tracking. Lead developer of DIRECT.
Tickmill Europe Ltd · Limassol, Cyprus
Deep learning and statistical models for financial time-series forecasting.
University of Amsterdam & Netherlands Cancer Institute
Supervisors: Jonas Teuwen, Jan Jakob Sonke, Clara I. Sánchez. Thesis: What We Do Sample, We Must Learn to Reconstruct — UvA DARE
Imperial College London
Thesis: Mirror Descent and Interacting Mirror Descent: Almost Dimension-Free Convex Optimization for Non-Euclidean Spaces
University of Cyprus · Top graduating student, Faculty of Pure and Applied Sciences
Thesis: Computational Approach of the Orr-Sommerfeld Equation with the Finite Elements Method (10/10)
University of Patras, Greece
Team fomofo, Foundation Models for Oncology Lab · All-data Track, linear probing
Marrakesh, Morocco
Vancouver, Canada
Calgary, Canada
Best AI Group Project, Imperial College London
Best academic performance, Department of Mathematics and Statistics
Faculty of Pure and Applied Sciences, University of Cyprus
Outside of work, I enjoy running, swimming, gym, skiing, pilates, cooking, reading fiction, and travelling. I thrive in collaborative settings and value multidisciplinary teamwork across research labs and hospitals.