Iain B. Styles

Quantitative analysis of multi-spectral fundus images

We have developed a new technique for extracting histological parameters from multi-spectral images of the ocular fundus. The new method uses a Monte Carlo simulation of the reflectance of the fundus to model how the spectral reflectance of the tissue varies with differing tissue histology. The model is parameterised by the concentrations of the five main absorbers found in the fundus: retinal haemoglobins, choroidal haemoglobins, choroidal melanin, RPE melanin and macular pigment. These parameters are shown to give rise to distinct variations in the tissue colouration. We use the results of the Monte Carlo simulations to construct an inverse model which maps tissue colouration onto the model parameters. This allows the concentration and distribution of the five main absorbers to be determined from suitable multi-spectral images. We propose the use of image quotients to allow this information to be extracted from uncalibrated image data. The filters used to acquire the images are selected to ensure a one-to-one mapping between model parameters and image quotients. To recover five model parameters uniquely, images must be acquired in six distinct spectral bands. Theoretical investigations suggest that retinal haemoglobins and macular pigment can be recovered with RMS errors of less than 10%. We present parametric maps showing the variation of these parameters across the posterior pole of the fundus. The results are in agreement with known tissue histology for normal healthy subjects. We also present an early result which suggests that, with further development, the technique could be used to successfully detect retinal haemorrhages.

Multispectral imaging of the ocular fundus using light emitting diode illumination

We present an imaging system based on light emitting diode (LED) illumination that produces multispectral optical images of the human ocular fundus. It uses a conventional fundus camera equipped with a high power LED light source and a highly sensitive electron-multiplying charge coupled device camera. It is able to take pictures at a series of wavelengths in rapid succession at short exposure times, thereby eliminating the image shift introduced by natural eye movements (saccades). In contrast with snapshot systems the images retain full spatial resolution. The system is not suitable for applications where the full spectral resolution is required as it uses discrete wavebands for illumination. This is not a problem in retinal imaging where the use of selected wavelengths is common. The modular nature of the light source allows new wavelengths to be introduced easily and at low cost. The use of wavelength-specific LEDs as a source is preferable to white light illumination and subsequent filtering of the remitted light as it minimizes the total light exposure of the subject. The system is controlled via a graphical user interface that enables flexible control of intensity, duration, and sequencing of sources in synchrony with the camera. Our initial experiments indicate that the system can acquire multispectral image sequences of the human retina at exposure times of 0.05 s in the range of 500-620 nm with mean signal to noise ratio of 17 dB (min 11, std 4.5), making it suitable for quantitative analysis with application to the diagnosis and screening of eye diseases such as diabetic retinopathy and age-related macular degeneration.

Multispectral imaging of the ocular fundus using LED illumination

We present preliminary data from an imaging system based on LED illumination for obtaining sequential multi-spectral optical images of the human ocular fundus. The system is capable of acquiring images at speeds of up to 20fps and we have demonstrated that the system is fast enough to allow images to be acquired with minimal inter-frame movement. Further improvements have been identified that will improve both imaging speed and image quality. The long-term goal is to use the system in conjunction with novel image analysis algorithms to extract chromophore concentrations from images of the ocular fundus, with a particular emphasis on age-related macular degeneration. The system has also found utility in fluorescence microscopy.

Selection of optimal filters for multispectral imaging.

Preece and Claridge [IEEE Trans. Pattern Anal. Mach. Intell. 26, 913 (2004)] have proposed a technique for selecting filters for the maximally accurate recovery of object parameters such as chromophore concentrations from a multispectral image of an object. Their selection criteria are derived from an analysis of a model of light propagation in the object and take into account both errors in the modeling process and errors in the image acquisition process, as well as the inherent behavior and structure of the model. We investigate their method on simulated image data and show that filters selected according to their criteria are demonstrably superior to other choices.

Model-based parameter recovery from uncalibrated optical images

We propose a novel method for quantitative interpretation of uncalibrated optical images which is derived explicitly from an analysis of the image formation model. Parameters characterising the tissue are recovered from images acquired using filters optimised to minimise the error. Preliminary results are shown for the skin, where the technique was successfully applied to aid the diagnosis and interpretation of non-melanocytic skin cancers and acne; and for the more challenging ocular fundus, for mapping of the pigment xanthophyll.

Quantitative interpretation of multi-spectral fundus images

Multi-spectral imaging of the ocular fundus suffers from three main problems: the image must be taken through an aperture (the pupil), meaning that the absolute light intensity at the fundus cannot be known; long acquisition times are not feasible due to patient discomfort; patient movement can lead to loss of image quality. These difficulties have meant that multi-spectral imaging of the fundus has not yet seen wide application. We have developed a new method for optimizing the multi-spectral imaging process which also allows us to derive semi-quantitative information about the structure and properties of the fundus. We acquire images in six visible spectral bands and use these to deduce the concentration and distribution of the known absorbing compounds in the fundus: blood haemoglobins in the retina and choroid, choroidal melanin, RPE melanin and xanthophyll. The optimisation process and parameter recovery uses a Monte Carlo model of the spectral reflectance of the fundus, parameterised by the concentrations of the absorbing compounds. The model is used to compute the accuracy with which the values of the model parameters can be deduced from an image. Filters are selected to minimise the error in the parameter recovery process. Theoretical investigations suggest that parameters can be recovered with RMS errors of less than 10%. When applied to images of normal subjects, the technique was able to successfully deduce the distribution of xanthophyll in the fundus. Further improvement of the model is required to allow the deduction of other model parameters from images.

Topological data analysis quantifies biological nano-structure from single molecule localization microscopy

The study of complex molecular organisation and nanostructure by localization based microscopy is limited by the available analysis tools. We present a segmentation protocol which, through the application of persistence based clustering, is capable of probing densely packed structures which vary in scale. An increase in segmentation performance over state-of-the-art methods is demonstrated. Moreover we employ persistence homology to move beyond clustering, and quantify the topological structure within data. This provides new information about the preserved shapes formed by molecular architecture. Our methods are flexible and we demonstrate this by applying them to receptor clustering in platelets, nuclear pore components and endocytic proteins. Both 2D and 3D implementations are provided within RSMLM, an R package for pointillist based analysis and batch processing of localization microscopy data.

Macular Pigment Quantification with Multispectral Retinal Image Analysis

Objective: Variations in macular pigment (MP) have been linked with changes in the risk of visual loss secondary n to macular pathology. MP density can be modified by diet; however there doesn’t appear to be a direct link between n dietary intake of MP components and MP density in the retina and clinicians therefore need a reliable and objective n method for MP measurement to establish if any intervention is yielding the required results. The objective of this n study was to investigate whether multispectral retinal image analysis (MRIA), a new technique for mapping retinal n pigments, is useful for measuring levels and distribution of MP to find differences between individuals with no clinical n evidence of macular pathology and those diagnosed with Age-related Macular Degeneration (AMD). nMethods: The study involved 90 volunteers from three subject groups: aged under 50 without AMD, aged 50 and n over without AMD and aged 50 and over with AMD. The experiments yielded 607 usable data sets that were used n for analysis. Multispectral image data was acquired at six selected wavelengths using a modified fundus camera. n MRIA maps of MP were computed from 3 × 3 mm regions of interest (approximately 10 degrees of visual angle) n centred at the fovea. Indices characterising MP distribution were computed both for individuals and for the three n subject groups. For comparison Macular Pigment Optical Density (MPOD) measurements were acquired from the n Macular Pigment Screener 9000 (MPS) based on the heterochromatic flicker photometry (HFP). Correlations for MP n quantities measured with the two methods were computed between MP quantity and age, MP quantity and AMD n diagnosis, and between the two methods. nResults: MP maps obtained from MRIA were consistent with known histology and in agreement with n expectations based on previous studies. Pooled results from the three groups suggest that the overall levels of MP n across both the fovea and the parafovea are on average higher in healthy under-50 individuals than that over-50 n with or without AMD. MP distribution might be more irregular in the over-50 groups than in the younger group. The n correlation between age and MP levels was weak as measured individually by both techniques. The MRIA indices n were not correlated with HFP-MPOD measurements for individuals, but high correlation was found between mean n HFP-MPOD and mean MRIA peak value for pooled results. nConclusion: MRIA has potential to offer an objective, fast and reliable method of measuring MP throughout the n posterior pole.

Spectrally constrained L1-norm improves quantitative accuracy of diffuse optical tomography

We consider L1-regularization of spectrally constrained DOT. Three popular algorithms are investigated: iteratively reweighted least square algorithm (IRLS), alternating directional method of multipliers (ADMM) and fast iterative shrinkage-thresholding algorithm (FISTA). We evaluate different regularizers and algorithms on a 3D simulated multi-spectral example.

Frequency domain 3D simplified spherical harmonics approximation: development, validation, and implication in bioluminescence imaging

A three dimensional (3D) photon transport model has been developed based on the frequency domain simplified spherical harmonics approximation (SPN) to the Radiative Transport Equation. Based on preliminary Monte Carlo studies, it is shown that for problems exhibiting strong absorption, the solutions using the 7th order SPN model (N = 7) are significantly more accurate than those from a standard Diffusion (SP1) based solver. This advance is of particular interest in the field of bioluminescent imaging where the peak emission of light emitting molecular markers are closer to the visible range (500 - 650 nm) corresponding to strong absorption due to hemoglobin.