New paper: Automatic identification of the temporal raphe from standard retinal scans

Glaucoma is a disease of the optic nerve. The nerve fibres that travel along the optic nerve to the brain are spread out across the retina, and can be visualised using an imaging technique called  optical coherence tomography (OCT). Individual nerve fibre bundles cannot be seen with standard clinical tests. It is typically assumed that if a nerve fibre bundle is lost in one part of the retina, then a person will have corresponding visual loss in that part of their field of view (i.e. visual field). Recently it has been recognised that different people can have quite different distributions of retinal nerve fibres, which means that the relationship between the location of loss of visual field and the location of nerve fibre damage differs between individuals and can be hard to predict.

Our lab has recently developed sophisticated models that overcome this problem to predict just how a particular nerve fibre pattern should map onto an individual patient’s visual field. These models depend critically upon a single measured parameter from the patient’s eye, namely the axis of symmetry between fibres that travel along the upper (superior) retina and those that travel along the lower (inferior) retina. This axis is known as the temporal nerve fibre raphe or temporal raphe for short (shown in the figure below as a red line).


Figure 1. The temporal raphe (red), where the nerve fibres travelling from the upper and lower retina meet, can be identified using retinal scans

The location of the temporal raphe is typically difficult to appreciate with standard imaging techniques. However, it has been shown recently that by collecting OCT data with scans of very high spatial density, the raphe can be visualised in exquisite detail and its orientation hence measured very precisely. Unfortunately, collection of such data requires a lot of time, good image quality, and particularly steady fixation, which are luxuries that we do not have in a clinical setting.

In this paper we set out to determine just how well the temporal raphe can be determined from standard scans acquired routinely in the clinic (‘macular cube’ scans). By acquiring additional high density scan data on the same eyes, we were able to measure how accurately several automated algorithms can estimate the orientation of the temporal raphe. Our results show that the best performing algorithm gave a mean error of 1.5°. The ability to quantify the orientation of the raphe with this level of precision paves the way for future work that uses these measurements to accurately map retinal structure (by OCT) to retinal function (by visual field), in order to more accurately assess the health of the retina and optic nerve in glaucoma.

This paper has been published in Biomedical Optics Express and is available to read here.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s