Trust but verify: Image-aware evaluation of radiology report generators
Evaluating generated radiology reports against the medical images they describe.
Associate Professor · Maître de conférences
Computer vision &
reliable medical AI.
I develop and evaluate machine learning methods for medical image analysis, with a focus on uncertainty, explainable AI, and learning with limited annotations.
PICSA team · Data Science Department · GIPSA-Lab
Principal Investigator (PI) · ANR JCJC EQUR-XSM
Selected publications
Publication list →VICCA: Visual interpretation and comprehension of chest X-ray anomalies in generated report without human feedback
Visual interpretation of chest X-ray anomalies described in generated reports.
IFSS-Net: Interactive Few-Shot Siamese Network for Faster Muscle Segmentation and Propagation in Volumetric Ultrasound
Muscle segmentation and propagation in volumetric ultrasound using few annotated slices.
Recent news
All news →Featured projects
All research →EQUR-XSM
Principal Investigator (PI)Evidential uncertainty and the reliability of saliency explanations in medical imaging.
SpinePredict
Scientific leadMultimodal AI for spine surgery planning, combining preoperative imaging, clinical notes and posture measurements.
SAFE-Med
Project leadSensitivity, fragility and explainability of medical imaging AI, in collaboration with UiT and Bio AI Lab.
Responsibilities & recognition
SICOM industry relations
Chargé des relations entreprises de la filière 3A SICOM (Phelma–Ense3)
Coordination of industry links and internships, alongside course responsibilities in computer vision and AI.
Teaching responsibilitiesPHAIMS
Physics-aware AI for imaging science, from space to atoms. The project combines AI with physical models for imaging applications in remote sensing, biomedical imaging and materials science.
About PHAIMSBIOINFORMATICS
Award for our co-authored work on domain-specific semantic similarity assessment of chest X-ray reports.
Award-winning publication