Modular discrete-wavelength light source synchronised with an intensity imaging camera for high-speed multispectral imaging

Colour plays an important role in the clinical diagnosis of many conditions. However, the receptors in the eye, as well as the sensors in a standard colour camera, provide only a limited representation of the visible spectrum in the form of three broad bands: red, green and blue. Multispectral images, where many narrow spectral bands are recorded, can yield information beyond what is possible by observation or conventional photography. In particular, advanced computational techniques can be used to derive biopsy-like quantitative maps characterising tissue composition in a totally non-invasive manner. Different sets of spectral bands can be selected, through optimisation techniques, to visualise different tissue components such as oxidised and non-oxidised blood, melanin, macular pigment, density of cells and organelles, etc. These techniques work well in situations where the tissue can be immobilised, but in some applications (e.g. retinal imaging, endoscopy), the tissue under examination is subject to involuntary movement and so can become displaced in a sequence of images. Whilst the displacement can be corrected using image processing techniques (image registration), the illumination of points also changes under movement, and this cannot be corrected by registration. It is thus necessary to obtain images at timescales faster than the natural movement of the tissue, at the same ensuring sufficiently long exposure times to maintain high image quality.A wide variety of commercial digital imaging systems are available based on broadband (e.g. white light) sources and digital video cameras. Where limited spectral bandwidth is required, narrow optical filters can be inserted. However, such systems may not allow a series of images at different spectral bands to be acquired with the speed sufficient for non-stationary objects. The principal objective of this project is to develop a modular imaging system which can be used in a range of imaging instruments to enable them to sequentially acquire digital images at an arbitrary number of wavelengths at video speed frame rates. The system will comprise a discrete multispectral light source synchronised with a scientific-grade monochrome imaging camera, connected via a computer controlled interface so that illumination and image capture are synchronised and the sequence and timing of the acquisition can be selected according to the needs of the application.The design will be exemplified by building and demonstrating a prototype system for use with a fundus camera for rapid multispectral imaging of the retina. Recent research has shown that quantitative maps of retinal blood derived from such image data may provide a substitute for invasive fluorescein angiography, commonly used to reveal retinal haemorrhages which are an early sign of diabetic retinopathy and may lead to blindness. Similar maps showing the quantity and distribution of Macular Pigment could be used to predict the risk of developing sight-threatening complications in age-related macular degeneration (ARMD).The proposed imager is a generic device with potentially broad applicability. As well as improving multispectral retinal imaging, the project will demonstrate the value of the imager as a flexible research device for exploratory studies in biosciences involving fluorescence microscopy studies of dynamic processes in living cells. Other applications envisaged include the use of the imager with diagnostic imaging systems such as endoscopes and optical tomography systems for non-invasive in-vivo diagnosis and monitoring of human tissues and organs.