Applications

Data fusion and registration are such important and widely used tools in medical imaging that it is not possible here to make an exhaustive review of their applications. Thus, we only present a few illustrative ones.
Registering time-series data from the same patient, in one or more modalities, allows to evaluate the progress of a disease. For instance, Ettinger [48] applies it to the follow-up of multiple-sclerosis patients. Wong [183] uses it in epilepsy diagnosis.
Registering images from the same patient in different modalities allow their simultaneous interpretation. For instance, Kapouleas [88] or Mangin [101] register functional Positron Emission Tomography (PET) with Magnetic Resonance (MR) images where the soft tissues' anatomy can be seen.
Registering medical images to the physical world permits their use in image-guided surgery. For instance, Davey [38] uses it for neuro-surgery planning; Herring [75] for image-guided surgery of the spine; Wasserman [178] for radio-therapy treatment planning.
Finally, registering images to a template anatomy can be useful for comparison purposes, to provide anatomical a priori to automatic segmentation procedures, ... For instance, Talairach [151] proposed a standard coordinate system for the brain. Kikinis [46] developed a digital brain atlas used - among others - by Warfield [176] in model-driven segmentation.