Under Review
Title withheld for double-blind review
International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI)
Academic Research
Developing accurate, efficient, and fair AI systems for medical imaging applications.
Focus Areas
Research Domains
Anatomical Landmark Detection
Accurate identification of anatomical keypoints in diagnostic images (MRI, X-ray), crucial for surgical planning, clinical support, and advanced image processing. Automation via deep learning enables fast, reliable analysis.
Label-Efficient Deep Learning
Due to privacy, expert-labeling needs, and time/cost constraints, annotated medical data are scarce. Using techniques such as Transfer Learning, Self-Supervised Learning, and generative models, the goal is to reach state-of-the-art results in low-data regimes.
Fairness & Debiasing in Medical AI
Investigating disparities in landmark detection across demographic attributes and exploring diffusion models (a class of generative image models) to mitigate hidden dataset biases—towards fairer, more reliable AI systems whose decisions are not driven by spurious correlations.
Research Collaborations
Academic partnerships and projects
Visiting PhD Research20 Sept. 2025 - 20 Dec. 2025
University of Oxford
Oxford, UK
DIBRIS, Department of Computer Science - University of Oxford
Department of Computer Science, University of Oxford — under Prof. Irina Voiculescu
Three-month visiting research period at the University of Oxford to expand research on landmark detection in the medical domain under Prof. Irina Voiculescu.
- Validated a cross-modality landmark detection pipeline for automated hip condition assessment on a matched cohort of 89 patients with paired MRI and X-ray images, achieving 87.5% diagnostic accuracy on real-world clinical data.
- Collaborated with clinical experts at the University of Oxford to translate research algorithms into practical diagnostic tools, while evaluating model adaptability using few-shot incremental landmark detection for unseen landmarks.
Supervisor
Prof. Irina Voiculescu
Duration
Three months
PhD Research ProjectNov. 2023 - Nov. 2026
Snoots S.r.l
Imperia, Italy
Master's ThesisMar. 2023 - Oct. 2023
Esaote S.p.A. Camelot Biomedical Systems
Genoa, Italy
Publications
8 peer-reviewed papers + 2 under review
2026
Under Review
Title withheld for double-blind review
BMC Bioinformatics
Conference
Automated Landmark Detection for Assessing Hip Conditions: A Cross-Modality Validation of MRI versus X-ray
IEEE International Symposium on Biomedical Imaging (ISBI)
2025
Conference
Diffusing DeBias: Synthetic Bias Amplification for Model Debiasing
NeurIPS — Conference on Neural Information Processing Systems
Workshop
Are X-ray landmark detection models fair? A preliminary assessment and mitigation strategy
International Conference on Computer Vision (ICCV) Workshop
Conference
A handful of data: Evaluating few-shot incremental landmark detection
International Conference on Image Analysis and Processing (ICIAP)
Journal
Artificial Intelligence in Chest Radiography — A Comparative Review of Human and Veterinary Medicine
Veterinary Sciences
Conference
Self-supervised pre-training with diffusion model for few-shot landmark detection in X-ray images
IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)
2024
Conference
Persistent Homology vs. Learning Methods: A Comparative Study in Limited Data Scenarios
STAG — Smart Tools and Applications in Graphics
Conference
Is in-domain data beneficial in transfer learning for landmark detection in X-ray images?
IEEE International Symposium on Biomedical Imaging (ISBI)
Academic Service
Peer review activities