MIDL: Medical Imaging with Deep Learning

Beschreibung

Imaging is a cornerstone of medicine, and deep learning has shown its potential to leverage the rapidly growing numbers of medical imaging studies. However, most deep learning research in computer vision and machine learning has focused on natural images. Investigating and deepening these techniques to the challenges of medical imaging is an important research challenge.

MIDL has a broad scope, including all areas of medical image analysis and computer-assisted intervention, where deep learning is a key element. Topics of interest include but are not limited to:

•  Semantic segmentation of medical images
•  Learning based image registration
•  Computer-aided detection and diagnosis
•  Image acquisition, reconstruction and synthesis
•  Transfer learning and domain adaptation
•  Learning with noisy labels and limited data
•  Unsupervised deep learning and representation learning
•  Uncertainty estimation for medical diagnosis
•  Interpretability and explainable deep learning
•  Integration of imaging and clinical data