EPSRC logo

Details of Grant 

EPSRC Reference: GR/H48170/01
Principal Investigator: Hawkes, Professor D
Other Investigators:
Colchester, Professor ACF
Researcher Co-Investigators:
Project Partners:
Department: Radiological Sciences
Organisation: United Medical & Dental Schools
Scheme: Standard Research (Pre-FEC)
Starts: 01 October 1992 Ends: 30 September 1995 Value (£): 170,920
EPSRC Research Topic Classifications:
Image & Vision Computing
EPSRC Industrial Sector Classifications:
Healthcare No relevance to Underpinning Sectors
Related Grants:
Panel History:  
Summary on Grant Application Form
To produce a system for accurate registration of image data from multiple modalities1. Based on a voxel representation of prior presence probabilities.2. Using different types of features to compute adequate starting estimates.Progress:We have shown how we can register MR and CT images of the base of the skull to an accuracy at least as good as interactive point based registration using anatomical knowledge of the expected distance between two (different) surfaces, the inner surface of the skull from CT and the outer surface of the brain from MR. We have also demonstrated how constraints provided by anatomical knowledge may be used to register images of blood vessels to images of the brain. The same methodology has been used to register 3D reconstructions of the skin surface, obtained from stereo video images after illumination with a structured light source, and 3D reconstruction of the same surface derived from MR images. This has been used in an application in image guided neurosurgery for the DTI funded project VISLAN (in collaboration with Roke Manor Research).Our most significant progress has been directed towards voxel based registration methods. The voxel intensity feature space produced by a pair of images from different modalities exhibit striking patterns specific to a particular application. These patterns change in a predictable way with mis-registration. We have devised and evaluated four different methods of this dispersion based on cross-correlation, entropy, variance of intensities, and moments of the histogram of the feature space. We have devised a multiresolution search strategy to determine the best estimate of the 3D rigid body registration transformation, which is both fast and robust. We have thoroughly tested the algorithm for MR and PET neurological image registration. This algorithm is now undertaking clinical trials in several centres. Unlike other approaches it is fully automatic and requires no pre-processing of the data and no user interaction. (To be presented at IPMI 95 and CAR 95). Work is in progress to evaluate the methods in other parts of the body and other modality combinations (in particular CT and MR). Accurate assessment of change over time is an important task in medical imaging. We have shown how accurate registration coupled with sinc interpolation during co-ordinate transformation can greatly improve assessment of very small changes in a patients status over time. We have developed a novel visualisation tool based on the volume rendering paradigm to display registered multi-modal volume data for planning and guiding surgery. Our methodology is finding application in planning and guiding skull base surgery and neurosurgery; functional imaging of the brain, including functional MR; the assessment of disease progression in multiple sclerosis; and precise location of tumour spread using PET and MR or CT. The work has been featured in the EPSRC/DTI publicity The Impact of IT, the BBC 2 programme The Net, SET 95 at the Science Museum and Science Now on Radio 4. We are in discussion with several companies regarding commercial exploitation of our work.
Key Findings
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Potential use in non-academic contexts
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Description This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Date Materialised
Sectors submitted by the Researcher
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Project URL:  
Further Information:  
Organisation Website: