Geometry Processing for Biomedical Imaging: 3D Model Representations and Multi-Modality

During the last decades, 3D medical imaging is enabling new and more accurate diagnosis and surgery inter- ventions, thanks to the development of new acquisition technologies and computerized methods of image analysis, modelling, reconstruction, segmentation and registration. Dedicated algorithms have been developed, aimed to ex- tract as much knowledge as possible from the images, with the ultimate goal of supporting and improving clinical practice, from diagnosis to surgery. This paper summaries the operations that can be performed on this kind of data, and strives to characterize the general work ow to transform raw data to useful 3D models. For each elaboration phase,the main existing methods are described, with particular focus on two active research areas: registration, in terms of integration and fusion of information acquired using either different modalities or in different times, and modelling, in terms of representing the object of interest in an efficient and lossless way