AMD Retinal Classification: Classification of Retinal Lesions Associated with Age-Related Macular Degeneration
The aim of the project is to use appropriate Artificial Intelligence (AI) algorithms in a software application that will aid in the clinical management of Age-related Macular Degeneration (AMD). AMD is the most important cause of blindness in the UK in older people and is estimated to account for the majority of people registered blind or partially sighted in the UK [1]. Clinical ophthalmology lags behind other medical fields such as radiology and oncology where image processing has been used to aid decision making for many years. This approach will help in the identification and classification of disease patterns with the aim of targeting specific treatments that can improve outcomes for patients suffering from choroidal neovascularisation (CNV) in AMD. The use of advanced signal processing, pattern recognition algorithms and decision software can aid the Ophthalmologist by allowing improved classification. This will in turn provide better prognostic indicators for disease by avoiding subjective processes that relies heavily on clinical expertise only.
The ophthalmologist takes many pictures from the back of the eye including colour photographs and greyscale images. The majority of eye clinics in UK and Ireland now use digital image capture and the images can be stored almost indefinitely. The doctor uses a specialised image capture technique called fluorescein angiography to examine the blood circulation of the retina. A patient receives a fluorescent dye that moves through the bloodstream and greyscale images are captured sequentially using a fluorescent camera. When the retina or its blood circulation is damaged, it shows up as bright fluorescent regions that are distinct from the rest of the image. Advances in image processing techniques [2] have produced many useful tools that have not yet been applied to clinical ophthalmology and that could help improve patient diagnosis and disease classification [3].
The project will investigate the use of complex image processing and artificial intelligence algorithms, which can provide more information about the fluorescent regions in the retinal images. The better these regions are characterised [4], the more accurate the classification of disease and therefore, the better the chances of successful treatments for patients. Fine-tuning of existing algorithms will provide better user interaction and custom functionality. Measurements can be taken over time and used to assess the effectiveness of therapeutic interventions.
This project will be of great interest to those who like the application of computer science to medicine.
Relevant Technological areas: Java, Java Advanced Imaging, Artificial Intelligence, Image Processing, Data processing.
- Klein R, Klein BE, Knudtson MD, Meuer SM, Swift M, Gangnon RE. (2007). "Fifteen-year cumulative incidence of age-related macular degeneration: the Beaver Dam Eye Study". Ophthalmology , Feb;114(2):253-62.
- Patton N, Aslam TM, MacGillivray T, Deary IJ, Dhillon B, Eikelboom RH, Yogesan K, Constable IJ, Retinal image analysis: concepts, applications and potential, Prog Retin Eye Res. 2006 Jan;25(1):99-127.
- Rapantzikos K, Zervakis M, Balas K., "Detection and segmentation of drusen deposits on human retina: potential in the diagnosis of age-related macular degeneration", Med Image Anal.2003 Mar; 7(1):95-108.
- Brankin E, McCullagh PJ, Black ND, Patton W, Mulgrew A, (Jun 2006) "The Optimisation of Thresholding Techniques for the Identification of Choroidal Neovascular Membranes in Exudative Age-Related Macular Degeneration", Proceedings of 19th IEEE Symposium on Computer Based Medical Systems, Utah, USA, 22-23 June 2006, IEEE,
Personnel Involved
First Supervisor: McCullagh, PJ Dr
Second Supervisor: Black, N Prof
Collaborator: William Patton
Collaboration: Dr William Patton (Opthalmology Reading Centre, Queen's University Belfast- domain expertise