Efficient 3D Corneal Modelling

Held By: Prof. Rami Qahwaji

Total Value:£270K (shared funding between University of Bradford and NHS National Innovation Centre)

Start date: 02/01/2013

End date: 02/01/2015

Introduction: The use of imaging techniques such as X-ray, MRI, and endoscopy, has become invaluable in the diagnosis and monitoring of disease. These technologies allow clinicians to see things that are either difficult to access, or are invisible to the naked eye, and new methods are constantly being sought to help with more and more medical areas.

Problem:Imaging of the cornea, the convex and transparent front-part of the eye which focuses light onto the retina, was not possible until the development confocal microscopy in the last decade or so. This technique captures high-resolution images of the cornea at different depths, and allows clinicians to diagnose and monitor injuries, dystrophies and diseases in the cornea that can potentially lead to blindness. The technique, while powerful, has been limited in its effectiveness. A single “corneal scan” can generate up to 700 images at different depths, which are often misaligned and blurry because of unavoidable eye-movements during the scanning process. After a scan it can typically take several weeks for the images to be analysed, in a process that is challenging for the clinician.

Project:Efficient 3D Corneal Modelling" project will use tools developed by researchers at the University of Bradford’s Centre for Visual Computing to overcome these problems, and revolutionise the use of confocal microscopy in corneal imaging. Prof. Rami Qahwaji, who leads the medical image processing group, has developed new techniques for the automatic, near-real time, analysis of confocal microscope images that convert the many 2D images into a single high-resolution 3D model. This 3D model is not only easier for the clinician to use, but it can be automatically analysed by computer algorithms to detect signs of early disease that are not visible to the human eye. By performing the analysis in near real-time, the results of the scan would be available to the clinician immediately after the confocal scanning procedure. This means that diagnosis can be made much more quickly, problems could be detected earlier, and the right treatments administered before serious disease develops.

Impact:The new technology is hoped to have two immediate effects: firstly by making the scanning process more accessible, it is hoped that more will be performed as part of routine assessment procedures, thus increasing the quality of patient assessments; secondly it is expected to reduce the necessity for corneal graft operations, which can be necessary when diseases have progressed too far. Reducing the need for graft operations will both benefit the patient, since the graft operations carry some risk, and will reduce the cost to the to the NHS, where over 2000 corneal grafts are performed every year. In addition the technology aims to make it possible to study the evolution of corneal structures over extended periods, which is important to study response to certain treatments, for example. It will also be used to gain a better understanding of common corneal pathologies, for example inflammation, early stages of corneal graft rejection and microscopic processes that can be sight threatening.

This research is supported by the NHS National Innovation Centre ( http://www.nic.nhs.uk/ ) and the University of Bradford.