Ординатура / Офтальмология / Английские материалы / Optical Coherence Tomography in Age-Related Macular Degeneration_Coscas_2009

The acquisition speed depends on the dimensions of the zone examined and especially on activation of the ART system

All images can be saved and shown in Heidelberg software

3on completion of the acquisition session.

The Heidelberg Eye Explorer software allows visualization of each image already obtained.

Opening the selected image, SLO and OCT images appear in display mode, with a green line (horizontal or vertical) on retinal imaging to identify the position of the OCT scan.

Clicking on this green line indicates and visualizes the OCT scan line by a calliper and visualizes the corresponding point on the OCT image and vice versa.

This feature is very useful to investigate tomographic characteristics of fine structures within the image.

Acquisition of a star with variable position, size, and density allows not only vertical and horizontal sections, but also oblique sections always centered on the fovea.

3D Visualization Mode

In 3D visualization mode, SLO and OCT images are overlapped to more clearly visualize the various scans and tomography position.

Images can be moved in three axes and zoomed in and out to more clearly visualize fine structures (Figure 3)

When displaying volume series, the 3D visualization mode can be used to analyze tomographs in three axes: x, y, and z, so that even C-scans (frontal or coronal plane) can be visualized.

Thickness Measurement Mode

In thickness measurement mode, the software defines the inner and outer limits of the retina, including the RPE, on OCT images.

For each position of the A-scan, retinal thickness is calculated and displayed on a graph (Figure 4)

The segmentation line can be displaced manually (inferiorly or superiorly) to improve the precision of retinal thickness measurements and to correct certain errors related to hyper-reflectivity or hypo-reflectivity not corresponding to the limiting membranes or retinal pigment epithelium (RPE).

The software elaborates a thickness (and volume) map based on the ETDRS chart, with the typical color scale (white and red for thicker areas to green and blue for thinner areas).

This mapping can be reproduced on corresponding fundus images with adjustment of its transparency, topography, and dimensions.

Follow-up

The Spectralis* apparatus uses an automatic eye-tracking system

Spectralis* can set scan line position automatically using the follow up feature.

The images acquired at the first examination can be used as reference images for subsequent follow-up examinations.

TruTrackTM

When the operator acquires follow-up images, Spectralis* is able to track eye movements and perfectly sets the OCT scan line at the same position on fundus images (either reference images or SLO fluorescein or ICG-A images). This system is called TruTrackTM

This could dramatically improve the capability of identification of minimal changes on consecutive examinations.

It allows evaluation of any changes due to therapy or to the natural course of various retinal diseases (Figure 5)

3SD-OCT images provide high-definition visualization of retinal layers and clear discrimination of the interfaces between different reflectance structures.

Noise reduction provides a marked increase in contrast with improved discrimination of fine structures.

Stratified retinal structures can be seen on TD-OCT, but the margins of layers with different reflectance characteristics are not clearly delineated, and it is difficult to identify interfaces between these elements.

Spectralis* images provide clear visualization and differentiation of each single layer, with sharper definition.

These images therefore show, starting from the surface:

▬The hyper-reflective points at the margin of the fovea and at the surface of the avascular clivus correspond to the adherence of the vitreous onto the inner limiting membrane (ILM). This hyper-reflectance is due to the perpendicularity of the single A-scan in relation to the vitreous. This is only visible with Spectralis* OCT.

▬The following five retinal layers (nerve fibers, ganglion cells, inner plexiform layer, inner nuclear layer, and outer plexiform layer) can be easily discriminated.

In particular, the margins of the nerve fiber layer are sharper, and its characteristics are more clearly visible.

In the ganglion cell layer, individual cells are almost visible, while Time-Domain OCT is unable to identify single hyper-reflective spots as corresponding to cells.

Intraretinal vessels are identified not simply as hyperreflective structures with a posterior shadowing effect into the retina, but with Spectralis* vessel shape is clearly shown, and it is often possible to identify the vessel wall and lumen.

The outer nuclear layer is visible as a hypo-reflectance structure very similar in TD-OCT and in SD-OCT.

The external limiting membrane can always be identified with Spectralis*, while it is often difficult to assess with TD-OCT.

The first hyper-reflective band underneath the outer nuclear layer corresponds to the inner/outer segment interface of photoreceptors.

With Spectralis*, it is possible to study photoreceptor layers in a incomparable way, in particular in dystrophies and in degenerative pathologies as dry and wet AMD. In these cases, the photoreceptor layer appears homogeneous with hyper-reflective fine deposits, and it is often possible to identify a jagged and irregular margin shape as an indirect sign of suffering.

The RPE layer is composed of two distinct hyper-re- flective bands separated by a fine hypo-reflective strip. This feature has already been observed with ultrahighresolution systems. They differ from results using normal high-resolution TD-OCT images in which the RPE layer appears as a unique hyper-reflective band.

The hyper-reflective outer band may correspond to RPE cells, but the origin of the inner band remains unclear.

Many studies suggest that this band corresponds to Verhoef’s membrane, constituted of tight junctions of RPE cells. According to another hypothesis, this band corresponds to basal infoldings and apical processes that enclose photoreceptor outer segments.

OCT images obtained with Spectral-Domain technology provide good visualization of the main vessels of the choroid, possibly due to the longer OCT wavelength and the decreased absorption by the RPE.

In addition, real-time averaging results in a sharper image with fewer artifacts derived from structures underneath the pigment epithelium.

Finally, the outermost layer visible on Spectralis* OCT is a moderately reflective structure that corresponds to sclera not detected by TD-OCT.

The Cirrus* apparatus is a Spectral-Domain OCT marketed by Carl Zeiss (Meditec, Dublin, CA).

This equipment constitutes a new generation after the Stratus* apparatus (the first and currently the most widely used Time-Domain OCT apparatus in clinical practice).

Like the other Spectral-Domain OCTs, the greatest improvement provided by Cirrus* is the increased acquisition speed, allowing very fast B-scan image acquisition and a larger number of A-scans for each single B-scan

(Figure 6)

The acquisition speed of the Cirrus* is about 100 times faster than that of the Stratus*.

Technical specifications of the Cirrus Zeiss* (Table 2).

Table 2. Cirrus* technical specifications.

The first mode consists of taking 5 consecutive raster horizontal scan lines in high-resolution definition. The operator can move the position of the five lines during the examination by analyzing SLO-red free fundus images.

Each raster scan line is constructed with single A-scans, which considerably increases the horizontal resolution.

In the other two modes, Cirrus* acquires retinal cube sections.

This cube is formed by 512 (vertical) x 128 (horizontal) A-scans for macular studies, or 200 x 200 A-scans for optic nerve head studies.

The acquisition time in each mode is about one second or a little more. This reduces the possibility of motion artifacts, making the presence of poor fixation–observed in many patients–less problematic.

Cirrus* has a powerful auto-regulation system for polarization and Z axis position, which facilitates acquisition. It generally provides very good examination quality, even for the higher bandwidth that reduces the importance of media opacity, such as in the case of cataract.

Wavelength

The OCT wavelength is slightly longer than that of the Stratus*, allowing greater penetration into the tissues and clearer visualization of subretinal elements.

Software and Image Acquisition

Raster Mode

In analysis mode, the operator can visualize all the different acquisitions. After raster acquisition, it is possible to examine each of the five consecutive B-scans, while clearly identifying the corresponding position on retinal imaging.

The position of the scan lines can be moved directly on retinal imaging and images can be zoomed on horizontal or vertical scans on the most relevant scan.

Cube Acquisition

The current software version allows OCT acquisition in 3 different modes

Post-processing of cubes is more versatile and current software now allows segmentation of the various layers to study any lesions.

In normal visualization, it is shown the retinal SLO image with vertical and horizontal scan lines and correspondent tomographies.

It is possible to move this position directly on retinal

3imaging or to zoom in horizontal or vertical tomography series and choose the most relevant image.

Retinal Thickness

Retinal thickness can be determined on a classical map based on ETDRS zones or on a graph (Figure 7).

A 3D map can also be obtained with white and red color coding for thicker areas and green and blue color coding for thinner areas.

Retinal Layers

The current software version is able to isolate the various layers of the retina. In particular, the profile of the RPE and the ILM are automatically identified and traced.

Analysis of retinal cube acquisition demonstrates the RPE and ILM, isolated from all other retinal layers, allowing separate visualization of either the single RPE or the ILM shape.

The software can also isolate RPE and ILM profiles and provide coronal (or frontal or en face) sections.

These sections are not only perpendicular to the laser beam, but also follow the shape of the retina facilitating quantification of a serous accumulation underneath the RPE or neuro-epithelium.

All cube acquisitions (512 x 128 or 200 x 200) can also be analyzed subsequently in advanced mode. The software simultaneously shows the tomographies in three axes (horizontal, vertical, and coronal) as well as the SLO retinal image.

Each single tomographic series can be analyzed more accurately, with a zooming effect and color coding or black and white visualization, adjusting contrast, and brightness.

By pointing and clicking on retinal images, the software shows the corresponding tomographies on vertical and horizontal axes.

Again, software can isolate RPE and ILM profiles, so that coronal tomographies visualization are not just perpendicular to laser direction, but they follow retinal shape making easier, for example, the quantification of a serous accumulation under RPE or under the neuro-epithelium

Follow-up

With the Cirrus* apparatus, it is possible to set the position of the cube on the reference examination as a basis for subsequent follow-up, with automatic repositioning of the cube on the same retinal area on each examination.

Advantages of the Cirrus*

The greatest improvement provided by the Cirrus* (and Spectral-Domain technology in general) is the high acquisition speed and, consequently, the greater number of A-scans for each B-scan. A very large volume of information is therefore acquired in a single examination.

In clinical practice, comparison between Cirrus* and Stratus* also demonstrates other major advantages.

Cirrus* allows ease of acquisition, even in the presence of media opacities. This feature is probably due to the higher wavelength used by Cirrus* allowing deeper penetration in the case of early cataract.

It is not uncommon to obtain good images with Cirrus* in patients who cannot be fully examined by Stratus* This is also the case in myopic patients, who are so difficult to examine (Figure 8)