Diana Mateus

Dr. Diana Mateus is a Full Professor (Professeur des universités) at Ecole Centrale Nantes, France, affiliated with the LS2N laboratory and the Signal Image et Son research team. Her primary research focuses on medical image analysis, machine learning, geometric computer vision, and shape analysis. She leads projects in computational sonography, surgical data science, and vision-language integration in medical contexts. Dr. Mateus has extensive experience in medical imaging applications, including PET-based lesion analysis, ultrasound segmentation, and deep learning for healthcare.

Her teaching includes courses on computer-aided medical procedures, medical augmented reality, and surgical robotics at TU Munich (noted in past sections). Research projects span computational surgineering, cross-modality image synthesis, and patient monitoring using wearable sensors.

Key contributions include developing IFSS-Net for muscle segmentation in ultrasound, curriculum learning for fracture classification, and graph-based frameworks for treatment response prediction in lymphoma. Her work bridges clinical needs with advanced AI techniques, addressing challenges in low-count PET imaging and weakly supervised learning.

Publications highlight innovations in medical imaging modalities (e.g., PET, ultrasound), with a focus on improving diagnostic accuracy and treatment planning. Collaborations include clinical institutions like the Deutsches Herzzentrum München (DHM) and teams at INRIA and MIT.