EQUR-XSM · PEOPLE
Team & contact
PRINCIPAL INVESTIGATOR
Dawood Al Chanti
Associate Professor · Grenoble INP – Phelma, UGA
Permanent researcher · PICSA team, Data Science Department, GIPSA-lab
His research combines computer vision and machine learning for medical-image understanding, focusing on anomaly detection, uncertainty-aware segmentation, multimodal fusion and explainable AI. As principal investigator, he coordinates EQUR-XSM and leads its two scientific work packages.
He earned an MSc in Machine Learning and Data Mining in 2016 through Université Jean Monnet and the University of Alicante, followed by a PhD in Image and Signal Processing at Université Grenoble Alpes in 2019. He conducted postdoctoral research in medical-image analysis at LS2N, Centrale Nantes, before joining Grenoble INP in 2021.
PHD RESEARCH · 2025–2028
Amrin Akter
PICSA team · GIPSA-lab · Grenoble INP – UGA
Toward Faithful and Accurate Post-hoc Visual Explanations in Medical Imaging
This PhD studies the instability and unreliability of post-hoc saliency methods through systematic comparison of faithfulness, localisation, consistency, sensitivity and stability.
The goal is an explanation method that is faithful by construction: attributions should reflect the features used by the classifier, without forcing emphasis on clinically relevant regions, such as tumours, that it ignored.
SCIENTIFIC COLLABORATOR
Mauro Dalla Mura
Associate Professor, HDR · Grenoble INP – UGA
Head of PICSA · GIPSA-lab
His expertise covers image and signal processing, computational imaging and multimodal data fusion, including remote-sensing applications.
He obtained a joint PhD from the University of Trento and the University of Iceland in 2011 and joined Grenoble INP in 2012.
Role in EQUR-XSM: contributes to WP1.2 and WP2, bringing expertise in uncertainty-aware explanations and multimodal fusion, particularly application A3.
GIPSA-lab profile ↗
SCIENTIFIC COLLABORATOR
Diana Mateus
Full Professor · Centrale Nantes
SIMS team · LS2N
Her research develops machine-learning methods for medical-image analysis, addressing volumetric and multimodal data, small or imbalanced datasets, and limited expert annotations.
Her work includes medical-image reconstruction and segmentation, and the analysis of PET images with clinical data.
Role in EQUR-XSM: contributes to uncertainty modelling and calibration in WP1.1, and to the segmentation and multimodal applications A2 and A3.
Researcher website ↗
Research roles follow the accepted ANR proposal. Current PhD membership and topic were supplied by the project coordinator. See also the GIPSA-lab project page.