Mathematical Optimization
Mathematical foundations of machine learning and imaging, with expertise in convex optimization, proximal algorithms, variational methods, inverse problems, and algorithmic optimization.
About
Medical AI and computational pathology
I am a Data Scientist at DEEMEA, where I design and validate AI-driven solutions for the structuring and analysis of medical data across multi-modal healthcare workflows. My work sits at the intersection of machine learning, medical imaging, and clinical data science where i turn complex biomedical data into interpretable, reliable, and actionable insights for research and patient care.
I am particularly interested in trustworthy AI systems that support collaborative, multi-centric clinical studies while ensuring privacy, robustness, and scientific rigor. My research and engineering work combine deep learning, optimization, and uncertainty-aware modeling to address real clinical needs.
Medical Image Analysis
Deep Learning
Computational pathology
Radiology
Few-Shot Learning
Convex Optimization
Clinical AI
Current role
Data Scientist, DEEMEA
Research focus
AI for medical image analysis and clinical decision support
Location
Paris, France
Research
Research Areas & Expertise
Medical Imaging & Radiology
Machine learning and deep learning for medical image analysis, with a focus on robust representation learning, image reconstruction, quantitative imaging, and computer-assisted diagnosis.
Computational Pathology
AI-driven analysis of histopathology and whole-slide images, including tissue and tumor characterization, architectural pattern recognition, quantitative biomarkers, and clinical outcome prediction.
Machine Learning & Deep Learning
Design and development of machine learning and deep learning models for biomedical applications, including representation learning, few-shot learning, weak supervision, and generalization under limited annotations.
Optimization-Driven Deep Learning
Integration of mathematical optimization and deep learning through algorithm unrolling, learned iterative schemes, and model-based architectures for interpretable and robust image processing.
Biomedical AI & Clinical Applications
Development of reliable AI methods for biomedical imaging, bridging mathematical modeling, computational imaging, and clinical challenges in diagnosis, prognosis, and quantitative pathology.
Doctoral work
PhD thesis
I completed my PhD, AI-Driven Methods in Histopathology for the Diagnosis and Prognosis of Hepatocellular Carcinoma, at CentraleSupélec, Université Paris-Saclay, in collaboration with the Center for Visual Computing Laboratory and Kremlin-Bicêtre Hospital. My work was supervised by Prof. Jean-Christophe Pesquet and Prof. Catherine Guettier, with additional supervision from Prof. Mounir Kaaniche and Prof. Amel Benazza-Benyahia.
Automated detection and classification of tumor architectures
Developed multi-scale, few-shot, and transductive deep learning frameworks to improve performance in histopathology under limited labels.
Recurrence risk modeling from WSI-derived features
Integrated AI-derived pathology phenotypes with clinical variables to support interpretable risk stratification.
Stain separation and normalization
Explored convex optimization and algorithm unrolling to improve stain consistency and feature stability in histopathological images.
Timeline
Recent milestones
Data Scientist at DEEMEA
Working on AI-driven medical data structuring and clinical AI systems for healthcare research and translational studies.
PhD defense
Defended doctoral work on AI for computational pathology and hepatocellular carcinoma diagnosis/prognosis.
Selected work
Publications
AI-Driven Methods in Histopathology for the Diagnosis and Prognosis of Hepatocellular Carcinoma
Thesis contribution
A Proximal Approach for Stain Separation and Normalization of Whole-Slide Histopathological Images
IEEE ISBI 2026
UNEM: UNrolled Generalized EM for transductive few-shot learning
CVPR 2025
Deep Learning Classification and Quantification of Pejorative and Nonpejorative Architectures in Resected Hepatocellular Carcinoma from Digital Histopathologic Images
The American Journal of Pathology
A transductive few-shot learning approach for classification of digital histopathological slides from liver cancer
IEEE ISBI 2024
Unrolled projected gradient algorithm for stain separation in digital histopathological images
IEEE ICIP 2024
Recognition
Achievements
🥇
1st place out of 256 teams in the DigiLut Data Challenge for Lung Transplant Rejection, organized by Hôpital Foch and Bpifrance. This work was carried out in collaboration with CentraleSupélec, Université Paris-Saclay, Inria, and OPIS.
Contact
Get in touch
I welcome opportunities in research collaboration, applied AI, and interdisciplinary work at the interface of healthcare, imaging, and data science.
Research profile
Clinical AI
Data Science
Computational Pathology
Medical Imaging