EPSRC logo

Details of Grant 

EPSRC Reference: GR/T02799/02
Title: Dynamic Image Registration Using Motion Models
Principal Investigator: Penney, Dr GP
Other Investigators:
Researcher Co-Investigators:
Project Partners:
Department: Medical Physics and Biomedical Eng
Organisation: UCL
Scheme: Advanced Fellowship (Pre-FEC)
Starts: 01 January 2005 Ends: 31 March 2007 Value (£): 222,773
EPSRC Research Topic Classifications:
Image & Vision Computing
EPSRC Industrial Sector Classifications:
Creative Industries
Related Grants:
Panel History:  
Summary on Grant Application Form
A modern radiology department has available an array of different imaging methods, which are used to help diagnose, plan and guide treatment. Each imaging method has its own set of advantages and disadvantages and because of this more than one imaging method is often employed to affect the treatment of a given patient. In some cases the different imaging methods provide complementary information about the patient. In other cases one imaging method may be used to diagnose and plan treatment, while another imaging method, due to its ease of use or real time nature, will be used to guide treatment during an intervention or surgery.Image registration is the process of aligning images so that corresponding features can be easily related. One aim when aligning images is to create a best of both worlds situation, where the best features from each image can be combined, so that the resultant image has greater value to the clinician than if each image was viewed separately. This can mean being able to compare the position of complementary information from two images, or it can mean aligning a high quality image taken before surgery with a lower quality, but real time , image acquired during surgery.Most of the established clinically used alignment techniques assume that the imaged anatomy is rigid. They are, therefore, limited to bony anatomy or regions enclosed by bone, such as neurological, skull base or orthopaedic applications. The extension of these techniques to allow for deformations has been the goal of a number ofusing motion models and by allowing alignment to dynamic sets of images such as ultrasound or fluoroscopy. A major problem with current techniques is that they can require a very long time to align images. This' precIides their use during a number of interventional or surgical procedures. This project aims to speed upthese algorithms by using prior information on how the anatomy can and cannot move and/or deform. This information will be provided by a motion model and it will be used to help guide the images into alignment, by ; restricting the size of the search which the algorithm must undertake and by guiding the algorithm to initially search for the most plausible alignments.At present most image alignment methods match two static images. However, dynamic sets of images frequently arise in interventional radiology, where the two most common image types are ultrasound and fluoroscopy. These are often favoured as the clinician can acquire images while manipulating needles or guidewires. This project aims to allow alignment methods to match on dynamic data sets, which should greatly increase the amount of image information available to the algorithm. A further aim is to investigate the ability of these algorithms to use information on how tissue moves to help align images. Currently it is the intensity information from both images which forms the basis upon which a good image alignment is judged. If motion information is available from both images then this could introduce extremely valuable information to aid image alignment, particularly for modalities such as ultrasound which have previously proved very difficult to align.This work will require investigations into the production of motion models. These models in themselves have the possibility to affect treatment: by allowing anatomical motion to be measured it may be possible to differentiate between healthy and diseased tissue, or to assess the success of some surgical procedures.The proposed work has the potential to make image alignment technology more applicable to a wide variety of clinical problems, not least in my particular interest area: the application of these techniques to interventional radiology.
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
Impacts
Description This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Summary
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: