Medical Image Computing

Medical Image Computing

We develop computational models and methods to extract relevant information from medical images.

Medical scanners routinely produce high-dimensional images containing millions of data points in a matter of minutes, highlighting various aspects of the structure and function of the living human body. Computation plays an increasingly central role in dealing with the resulting torrent of data, as the traditionally-used visual eyeballing of medical scans by human experts is extremely tedious; easily misses subtle disease progression over time; and lacks reproducibility. By automatically extracting precise quantitative information from the images, computation has enormous potential value in early detection/diagnosis, therapy planning, treatment evaluation, and in the development of new therapies and clinical research.

The group maintains a strong collaboration with the Athinoula A. Martinos Center for Biomedical Imaging at Massachusetts General Hospital, and actively contributes to the development of the open-source neuroimage analysis software suite FreeSurfer. We are also the coordinator of the Translational Brain Imaging Training Network (TRABIT), an EU-funded consortium focused on bringing computational neuroimaging methods into the clinic.


Koen Van Leemput
Groupleader, Professor
DTU Health Tech
+45 45 25 52 95