Diabetic Eye Screening Services in Scotland: A Training handbook – July 2003: page 23

[Previous] [Contents] [Next]

INDOCYANINE GREEN ANGIOGRAPHY

Indocyanine green angiography (ICG) records the circulation dynamics of the choroid. Dye is injected intravenously in the arm and photographs are taken using infra-red illumination. Infra-red light is efficient at penetrating the pigmented layers of the retina, exudates, thin layers of subretinal blood and some media opacities to show the choroidal vasculature more clearly than fluorescein angiography.

INDOCYANINE GREEN ANGIOGRAPHY photo

Idiopathic choroidal polypoidal vasculopathy. Top left- red-free image showing pale lesion adjacent to disc. Top right- fluorescein angiogram shows non specific hypofluorescent lesion with non-specific hyperfluorescence. Bottom-left- early ICG image showing choroidal vessels. Bottom-right- late ICG image showing vascular choroidal polyps.

Non-Mydriatic Cameras

Non-Mydriatic Cameras photo

Non-mydriatic fundus cameras work without the need to administer drops into the patient’s eyes as long as the patient’s pupils are 4mm in diameter. This camera operates in a similar fashion to the mydriatic camera, however, rather than a visible halogen viewing lamp, the viewing light is derived from an infra-red source.

The infra-red illuminating source then passes through a series of mirrors, into the eye and the resultant image is viewed on a monitor, which displays an image of the fundus so that it can be focussed, and to ensure that the area of interest is in view.

When the trigger is fired, an electronic flash is released simultaneously so that a conventional photograph can be taken. The procedure requires the patient to sit in a darkened room for approximately 3-4 minutes where the eyes will naturally dilate. As soon as one photograph is taken, the pupils will immediately constrict; however, they will physiologically re-dilate after approximately another 3-4 minutes. Then a picture, of the second eye can be taken.

CAPTURE MEDIA

Fundus cameras use different types of photographic media to record images:

  • 35mm film
  • Polaroid film
  • Digital imaging

35mm film is used in fundus work mainly to record the fundi for teaching purposes, as the resolution of films is of a high standard.

Polaroid film has been used in the past to record images for screening purposes such as for diabetic retinopathy; however, this has been superseded by digital imaging.

Digital Imaging

Interest in digital imaging has grown rapidly in recent years as technology has improved and its cost has fallen. Besides medical imaging, digital imaging is now also widely used in life science image analysis, industrial vision, satellite remote sensing, film special effects, criminology, high energy physics and military guidance systems. The increasing popularity of digital photography has brought digital imaging to the masses. Digital imaging is fast replacing traditional photographic film for more and more applications.

Digital images of the retina represent the intensity of light reflected from the retina and its spatial arrangement as an array of numbers that can be stored on a computer. Each number makes up a single picture element known as a pixel. Digital images may be copied any number of times and transferred anywhere in the world without any loss in quality whatsoever.

True digital images can only be obtained using a charged coupled device. These images can have a variable resolution. A resolution of at least 1365 x 1000 has been recommended by the National Screening Committee and the Health Technology Board for Scotland. This is to enable the detection of microaneurysms, which are small circular lesions, the earliest visible feature of diabetic retinopathy. This level of resolution is also required for automated grading using computer software.

However, the higher the resolution, the larger the size of the image file:

  • Black and White
    • 1024 x 1024 pixels per image (1MB)
  • Colour
    • 512 x 512 pixels (1.5MB)
    • 728 x 728 pixels (3MB)
    • 3000 x 3000 pixels per image (15 MB)

A computer with larger capacity RAM and a faster processor for capture and storage is then required.

IMAGE ACQUISITION

Digital imaging may be used for all kinds of retinal photography, including red-free, colour and fluorescein angiography. The photographs are available almost immediately after acquisition, making it ideal for use in a screening situation.

Digital images may be acquired directly using a digital camera attached to the fundus camera or a video camera and computer video capture card, or indirectly using a slide scanner. The majority of image capture devices are based on CCDs (Charge-Coupled Devices), which have light sensitive integrated circuits (“chips”).

Some video cameras acquire the image using a system known as interlaced scanning, where the odd rows of the image are stored first, followed by the even rows. This is how ordinary television cameras acquire images. However the system can be problematical when acquiring retinal images as the short duration of the fundus camera flash means only the odd or even lines are correctly illuminated, giving a striped appearance to the images. The popular Sony DXC950P video camera uses interlaced scanning. Modern high resolution cameras generally acquire the entire image in one go rather than using interlacing.

The eye contains only three types of colour receptors (cone cells). As a result all visible colours may be produced by mixing a maximum of three primary colours. This is why computer monitors and television screens are able to form colour images using a pattern of red, green and blue dots. Likewise colour cameras work by capturing three images (a red image, a green image and a blue image) which when combined represent the colour scene.

IMAGE ACQUISITION photos

Note that the red and green channels are much brighter than the blue channel since little blue light is reflected by the retina. The green channel is equivalent to the normal red-free image. Colour images are therefore usually three times larger than a monochrome image with the same number of pixels. Colour images may be acquired in two ways:

THREE CHIP DEVICES

These devices contain three CCD detectors. The incoming light is divided in three. One copy is passed through a red filter before being directed to one of the CCDs, another copy is passed through a green filter and the remaining copy through a blue filter. Each of the three colour channels has the same resolution. Examples of three chip cameras are the Sony DXC950P and the JVC KY-70U.

SINGLE CHIP DEVICES

Tiny coloured filters are placed over each individual light receptor on the CCD chip, making each pixel sensitive to either red, green or blue light. Usually the coloured filters are arranged in a pattern with twice as many green sensors as red or blue sensors because the eye is most sensitive to changes in the green light. The final image is created by estimating the values for the two missing colours from the values of the surrounding pixels. Consequently for the same number of pixels, the true resolution of a three chip camera is slightly higher than that of a single chip camera. Examples of single chip cameras include the Canon D30 and D60, and the Nikon D1.

[Previous] [Contents] [Next]