Restricted Access

You must be logged in to view this content.

Channel surfing: Fundus photography in a new light

$currentPage/@nodeName

Roman Serebrianik
BOptom PGDipAdvClinOptom PGCertOcTher FACO
Lead Optometrist Primary Care, Australian College of Optometry

 

Digital ocular fundus photography is a common and widely available imaging modality in optometric practice. While most practitioners are familiar with colour and red-free fundus photography for analysing vascular structures in the retina, a considerable amount of additional diagnostic information can be gleaned from exploring other ways of viewing the retinal photos captured by a fundus camera.

Monochromatic or split-channel viewing

Digital fundus cameras capture images of the fundus with a white light flash. Contained in that light are red, green and blue colour channels which when combined, render a full-colour image of the retina. With the use of standard image analysis software, practitioners can `split' the RGB light channels and view fundus images with each colour. Essentially, this simulates a filter blocking the transmission of wavelength bandwidths corresponding to the other two colours. It's important to note that objects appear lighter (reduced contrast) when viewed through a light channel of the same colour.

Figure 1 shows a composite (white colour) fundus image and the same image viewed through individual red, green and blue channels. Note that different features of the fundus are accentuated in different light channels due to differing absorption spectra of retinal luteal pigment (most densely concentrated at the macula), haemoglobin and melanin. In other words, by eliminating channels, the visibility of various structures can be enhanced. Haemoglobin in blood vessels is best seen in the green channel. Melanin in RPE/choroid is best seen in the red channel. Features of the retinal architecture are best examined in the blue channel. Disregard a camera lens artefact at 2 o'clock position.

919 Figure 1 - F

Figure 1. The composite fundus image (top left) and the same
image viewed through the red, green and blue channels.

Blue channel

Blue channel (B) increases the visibility of more anterior retinal structures, for example, the internal limiting membrane and nerve fibre layer, which are virtually transparent in white light and likely to be `drowned out' by the deeper layers of the retina and the RPE.

Therefore, this single channel (approximately 490-510 nm) is of particular benefit when viewing xanthophyll pigment concentrations in photoreceptor layer (lutein, zeaxanthin and meso-zeaxanthin), epiretinal membranes, macular holes or retinal nerve fibre layer defects in early glaucoma. While there are currently no established clinical protocols for correlating retinal pathology and xanthophyll appearance on fundus photography, the blue channel does allow a better view of the macula pigment compared to white light. Similarly, this channel would also be useful in differentiating small cotton wool spots (retinal nerve fibre layer infarct) from a druse.

It is worthwhile to note that since short-wavelength light is more prone to scatter, the quality of the blue channel is particularly susceptible to degradation due to cataract or corneal opacification.

Figure 2 compares full colour and blue channel view of the same fundus in a patient with moderately advanced glaucoma. Note the emphasised visualisation of the superior nerve fibre layer losses in the blue channel. A flash artefact is also present next to the disc.

919 Figure 2 - F

Figure 2. A patient with moderately advanced glaucoma. The blue
channel emphasises superior nerve fibre layer losses.

Green channel

Green channel (G) bandwidth is in the middle of the visual spectrum (530-550 nm) and is less affected by cloudy media. The green channel provides great definition to the outer retinal architecture. Haemoglobin in the blood readily absorbs light in this wavelength region, depending on the level of oxygenation, so this channel is particularly useful for documenting retinopathy, anomalous blood vessel calibre changes and other subtle vasculopathies. Drusen and exudates will also appear enhanced in this view.

Isolating the green light channel produces an image similar to the familiar red-free fundus image; however, traditional red-free photography uses a broadband filter to block the longer wavelengths, while allowing mixed short (blue) and medium (green) wavelength bands to be viewed together.

Figure 3 shows the fundus of a patient with diabetic retinopathy. Note the enhanced contrast of haemorrhages in the green channel due to absorption by haemoglobin (arrows). Also note the evident venous dilation and some areas of beading (X).

919 Figure 3 - F

Figure 3. A patient with diabetic retinopathy: green channel enhances
contrast of haemorrhages, venous dilation and beading

Red channel

The red (R) channel uses longer wavelength light, with the peak around 590-620 nm. Blood components and retinal vessel walls look indistinct in this channel, as does the optic nerve head. However, retinal pigment epithelium (RPE), choroidal nevi and other pigmented features will stand out better against the paler fundus appearance, and subtle atrophic macular degenerative changes or choroidal melanoma features may become more visible than they would be in white light.

The red channel highlights the layout of the choroidal layer. Finer details of the choroidal layer may be indistinct when viewed through the overlying retina in white light.

Figure 4 compares full colour, red channel, green channel and blue channel photography of a choroidal nevus. Note the nevus has the highest contrast in the red channel (due to the abundance of melanin). There is significantly less visibility in the green channel and none in the blue channel.

919 Figure 4 - F

Figure 4. The choroidal nevus has the highest contrast in
the red channel, but is invisible in the blue channel

Additionally, the green channel highlights the presence of a microaneurysm at the superotemporal edge of the fovea, a feature absent in the red channel view.

The benefits of composite diagnoses

Single channel viewing can be particularly useful in observing multiple features of the fundus, which may be difficult or challenging to view together in an overlapping full colour image. In a particularly apt example (Figure 5), the patient's retina exhibits:

  • Geographic atrophy due to atrophic macular degeneration—best appreciated in the red channel, which allows ready visualisation of choroidal vessels through the missing RPE
  • Multiple small choroidal naevi and peripapillary atrophy—again, best viewed in the red channel
  • A small intraretinal haemorrhage adjacent to the macula—best viewed in the green channel
  • Disrupted macular pigmentation in the photoreceptors of the macula—seen best in the blue channel.

Thus a complex clinical presentation and image may be `split' into composite diagnoses by the depth of pathology in the photographed tissue.

Figure 5 shows a case of multiple retinal pathologies, presented in both colour and individual red, green and blue channels.

919 Figure 5 - F

Figure 5. A case of multiple retinal pathologies, revealed through channel separation

It is important to remember that both normal and abnormal fundus colouration is highly variable in all patients and depends on many factors, such as the patient's age, ethnicity, media and refractive status, so that the appearance of retinal structures in the RGB channels may vary.

Fundus photography is a valuable imaging modality in clinical practice, and should be used alongside other diagnostic tools such as ocular coherence tomography and fundus autoflorescence (FAF) in the management of ocular disease. It is my hope that this brief article encourages the reader to look at retinal fundus photography in a new or at least not exclusively `white' light in the management of their patients and ocular disease.

Acknowledgement

The author acknowledges the assistance of Dr Adrian Bruce in preparing this article.

All images were taken with a Canon CR-1 digital fundus camera.



Like us on Facebook




Subscribe to our News RSS Feed

Latest Tweets




Recent Comments