Stage Ingénieur R&D – "Anatomy Score Maps for Neurovascular X- ray" (H/F)
Core
Develop and compare deep learning methods to generate pixel-level maps highlighting artificial structures (catheters, stents, etc.) in neurovascular X-ray images.
Role type
Research intern (Master 2) in medical imaging and deep learning
Builds
A functional prototype generating artificial structure maps for neurovascular X-ray images
Domain
Medical imaging / Deep learning / Computer Vision
Deliverable
research
Required skills
Deep learning (CNN, U-Net), Computer Vision, Image Processing, Python, PyTorch/TensorFlow/Keras, Statistics, Machine Learning
Preferred skills
Self-supervised learning, Anomaly detection, Medical domain knowledge
Technologies
U-Net, SimCLR, BYOL, PatchCore, PaDiM, Deep SVDD, Normalizing flows
Responsibilities
Review state-of-the-art in segmentation and anomaly detection; Implement and train deep learning models; Evaluate method robustness against unknown devices; Define evaluation protocols; Develop a functional prototype.
Seniority
Intern (Master 2 level)