TOSBI – Towards Scalable and Generalized Solutions in Biomedical Images (2025–2028)
The TOSBI project — Towards Scalable and Generalized Solutions in Biomedical Images — aims to create deep learning methodologies that scale to large bioimage datasets and generalize across imaging modalities, laboratories and experimental settings. The project emphasizes annotation-efficient learning and the adaptation of foundation models to biomedical imaging tasks.
Led by the Computer Vision and Pattern Discovery (CVPD) group at the University of the Basque Country (UPV/EHU), TOSBI addresses central challenges in modern biomedical image analysis: limited high-quality annotations, the computational cost of state-of-the-art models, and the difficulty of achieving robust cross-domain generalization.
🔬 Scientific Goals
- Develop and benchmark annotation-efficient learning strategies (self-supervised, semi-supervised, few-shot and zero-shot) tailored to biomedical imaging.
- Improve scalability of training and inference workflows for large microscopy and histopathology datasets.
- Create domain adaptation and domain generalization methods enabling models to transfer across imaging modalities, laboratories or acquisition conditions.
- Adapt and validate foundation models for downstream bioimage tasks (segmentation, detection, classification), providing practical tools for the research community.
- Release open and reproducible software components and evaluation pipelines fully compatible with community platforms such as BiaPy.
👥 Project Team
Principal Investigators
- Dr. Fadi Dornaika (Ikerbasque Research Professor, UPV/EHU)
- Dr. Ignacio Arganda-Carreras (Ikerbasque Research Associate Professor, UPV/EHU; DIPC; Biofisika Institute)
Research Team
- Dr. Urtzi Ayesta (Ikerbasque Research Professor, UPV/EHU)
- Dr. Nagore Barrena (Assistant Professor, UPV/EHU)
- Dr. Unai Elordi (Assistant Professor, UPV/EHU)
Work Team / Collaborators
- Aitor González-Marfil (PhD student, CVPD)
- Estibaliz Gómez de Mariscal (Postdoctoral researcher, Portugal)
- Daniel Franco-Barranco (Postdoctoral researcher / collaborator, UK)
- Dr. Jinan Charafeddine (De Vinci Higher Education, France)
- Dr. Fares Bougourzi (University of Salento, Italy)
- Dr. Donglai Wei (Boston College, USA)
- Software Engineer (SE) — to be hired (WP8)
- PhD Candidate (FPI) — to be recruited (linked FPI contract)
🕒 Timeline
- Start date: 1 September 2025
- End date: 31 August 2028
💰 Funding
- Funded by the Spanish Ministry of Science, Innovation and Universities (MCIU) through the Generación de Conocimiento 2024 (Proyectos de Investigación No Orientada) program (MICIU/AEI / 10.13039/501100011033 and FEDER, EU).
- Total project funding: 143,250 € (including indirect costs)
- Reference code: PID2024-157485NB-I00
📄 Project Documentation
- Related post: PhD announcement
📬 Contacts
For project enquiries, please contact:
Ignacio Arganda-Carreras
Ikerbasque Research Associate Professor
My research interests include image processing, computer vision, and deep learning for biomedical applications.
Fadi Dornaika
Ikerbasque Research Professor
Ikerbasque Research Professor with expertise in computer vision, machine learning, and pattern recognition.
Nagore Barrena
Assistant Professor
My research focuses on computer vision and deep learning, with applications to biomedical and natural images.
Unai Elordi
Assistant Professor
My research interests include computer vision, pattern recognition, and artificial intelligence for intelligent video analytics.
Aitor González-Marfil
PhD Student
My research focuses on generalizable deep learning methods for computer vision.
Daniel Franco-Barranco
Postdoctoral Researcher
My primary focus is on the development of deep learning solutions for the segmentation of organelles in large-scale and multimodal electron microscopy images.