The unique properties of the confocal Scanning Laser Ophthalmoscope [cSLO] have been exploited in tandem with computer-aided image analysis to detect and quantify macular lesions. Healthy volunteers and patients with macular ischaemia, oedema and hole formation underwent fluorescein angiography, tomography and indirect mode imaging using the cSLO. For macular ischaemia, a computer program has been developed to identify poorly perfused areas based on fluorescence and temporal data extracted systematically from fluorescein angiograms. These parametric images, <I>viz</I>., <I>maximum fluorescence, time-to-maximum </I>and <I>ischaemia</I>, combine physiological with anatomical data. The ischaemia program has been validated using a human Gold Standard on a panel of five angiograms using Receiver Operator Characteristic analysis. The program performs favourably against a panel of ophthalmologists for angiograms of diabetic retinopathy. Computer algorithms have been developed for extracting topographical features from tomographic images of macular lesions causing thickening or thinning, such as oedema and holes. The information has been presented as <I>topography </I>and <I>thickness</I> parametric images. The novel <I>thickness</I> images show three-dimensional data in an easily assimilable form. The indirect mode technique readily reveals retinal and subretinal lesions, including macular oedema, macular hole, choroidal naevi and retinal pigment epithelial detachment without the need for any specialised viewing equipment or image processing. We have made significant progress towards developing a research and potential clinical tool for the detection and accurate quantification of macular pathology to facilitate the better understanding, diagnosis and clinical management of these blinding conditions.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:367438 |
| Date | January 2000 |
| Creators | Lim, Chea Siang |
| Publisher | University of Aberdeen |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.0025 seconds