Our thesis is that fusion/registration of magnetic resonance and
ultrasound information can make significant contributions to the initial
diagnosis and localisation of prostate cancer, and as such to the staging
process (surgery/radiotherapy). In addition, the available information can
be used to improve localisation of biopsies and therapy (both the
targeting of specific areas and the monitoring of treatment).

Current approaches to the fusion of MRI and ultrasound prostate
information tend to rely on manual intervention or tracking systems. We
propose to develop a data driven approach to this fusion/registration
problem, which will use prostate boundary and internal anatomical region
information. The automatic segmentation of these prostate features will be
based on shape and texture analysis techniques. The resulting fusion of
MRI and ultrasound data will form the basis for the clinical evaluation
with respect to the diagnosis and localisation of prostate cancer