Image segmentation and registration

Motivation

Medical images capture anatomical and functional information of our heart, and image analysis solutions are provided by companies to extract relevant metrics. The two main analysis tasks are to segment (contour the anatomical structures) and to register (spatially align structures and retrieve deformation fields) images, and these are two pre-processing tasks that are needed to personalise computational models.

Our approach

CMIB researchers investigate on the synergies between image analysis and cardiac modelling in order to improve the automation, accuracy and robustness of segmentation and registration tasks. Image acquisition and analysis artefacts can be corrected (regularised) through our understanding of anatomy and physiology.

Further reading