OCT Image Analysis in Glaucoma

Cirrus RNFL Analysis - Orientation

This page will refer to many different sections of the RNFL analysis. Key components of this printout are below:
1. Signal strength
2. Horizontal tomogram identifying the edge of both the cup (red) and disc (black)
3. Circular tomogram identifying the inner and outer border of the RNFL (shows instrument segmentation of this layer)
4. RNFL thickness (heat) map
5. RNFL deviation map
6. RNFL global parameters
7. TSNIT curve
8. RNFL quadrant analysis
9. RNFL clock hour analysis

Cirrus GCA Analysis - Orientation

This page will refer to many different sections of the GCA analysis. Key components of this printout are below:
1. Signal strength
2. Horizontal B-scan identifying the inner and outer borders of the ganglion cell-inner plexiform layer (shows the instruments segmentation of this layer)
3. Thickness (heat) map
4. Deviation map
5. GCA global parameters
6. GCA sector analysis

Cirrus RNFL Analysis - Signal strength

A motion artefact: RNFL heat map (top left), deviation map (top right) and circular tomogram (bottom)

Motion artefacts occur due to the movement during scan acquisition, most commonly due to eye movement. Motion artefacts are most easily visualised as a break or discontinuity in the retinal vasculature on the deviation maps. The deviation map may consequently flag regions of apparent thinning

If the RNFL scan circle is affected, the retinal profile can be affected on the circular tomogram.

Blink artefact: RNFL heat map (top left), deviation map (top right) and circular tomogram (bottom)

Blink artefacts are caused by a patient blinking during acquisition and appears as black bands of missing data which span the entire maps. This may affect disc and/or cup delineation.

If the blink artefact affects the scan circle, vertical blanks of missing data can appear on the circular RNFL tomogram. This will affect the instruments measurement of the RNFL thickness.

Media opacity: RNFL heat map (top left), deviation map (top right) and circular tomogram (bottom)

Media opacities such as floaters can appear as focal data gaps (black areas) on the thickness map. This can subsequently results in apparent thinning (red pixels) on the deviation map.

Opacities affecting the scan circle manifest as vertical black shadows interrupting the retinal profile and RNFL segmentation.

Vignetting: RNFL heat map (top left), deviation map (top right) and circular tomogram (bottom)

These arcuate areas of missing data are typically due to shadowing from the pupil margin. The circular tomogram will show a degrading scan signal in the affected area (white arrow) of the scan.

RNFL vertical tomogram (1), RNFL circular tomogram (2) and Left horizontal GCA B-scan (3)

Image 1 shows identification of the edge of the optic cup (red arrow) and optic disc (white arrow).

Image 2 shows accurate delineation of the RNFL on the CIrrus RNFL analysis.

Image 3 shows accurate delineation of the GCL+IPL on the Cirrus GCA analysis.

RNFL segmentation error (RNFL heat map, deviation map and circular tomogram)

The RNFL in the tomogram is incorrectly delineated. This can be due to errors by the instrument’s algorithm or the presence of pathology e.g. floaters

Inaccurate optic disc and cup margin delineation (RNFL heat map, deviation map and circular tomogram)

Depiction of the cup (red arrow) and disc (white arrow) margins are incorrectly delineated. Inaccurate delineation of Bruch’s membrane opening is often associated with the presence of PPA.

A "normal" RNFL thickness map (Cirrus OCT)

RNFL thickness map showing superior and inferior thinning

RNFL Deviation Map in a "normal" subject (Cirrus OCT)

RNFL Deviation Map showing superior and inferior thinning

RNFL Analysis Global parameters (Cirrus OCT)

RNFL Analysis Global Parameters in a patient with glaucomatous RNFL loss in the right eye

The heat maps show RNFL thinning superiorly and inferiorly in the right eye. Although the average RNFL thickness is within the normative database, it is important to notice that the RNFL thickness is thinner in the right eye compared to the left eye. Disc area (reflecting disc size) is equal between the two eyes and reflects that these are average-sized discs.

Example of inaccurate optic disc and cup margin delineation affecting global parameters

The global parameters here indicate that the left disc is significantly larger than the right. However, when we look at the horizontal tomograph, we can see that the edge of the cup and disc have been misidentified in the left eye. This will subsequently affect other disc parameters including disc area, rim area and cup/disc ratio.

TSNIT curve (Cirrus OCT) in a normal eye

Note the symmetry between the two eyes in the TSNIT curve. Both curves fall within the green area of the curve indicating the RNFL thickness values are within normallimits.

A TSNIT curve showing a focal RNFL loss superiorly and inferiorly in the right eye

This TSNIT curve shows localised RNFL thinning in the right eye compared with that of the left eye (locations are indicated by the blue arrows). Although the curve for the right eye still falls within the normative distribution, this asymmetry indicates probable superior and inferior RNFL loss in the right eye.

Normal RNFL quadrant and clock hour thickness(right eye)

The RNFL quadrant and clock hour analysis shows RNFL thickness values to be within the normative distribution in all quadrants and clock hours.

Quadrant and clock hour analysis showing a reduced RNFL thickness superiorly and inferiorly

The RNFL quadrant and clock hour analysis shows RNFL thinning superiorly and inferiorly

RNFL quadrant, clock hour analysis and circular tomogram

It is important to note that the distribution of blood vessels can cause normal variations in RNFL thickness. Accurate segmentation of the RNFL is critical to check to ensure these measurements are accurate in each eye.

The RNFL thickness values in this patient are relatively symmetrical between the two eyes, except for the clock hours indicated by the red squares. The RNFL at 2 o'clock in the right eye is flagged as being thicker than the normative distribution, and is thicker than the corresponding clock our in the left eye.

Examining the associated circular tomogram, we can see that the RNFL is thickened in the area indicated by the yellow arrow. On closer inspection, you can make out a blood vessel at this point (indicated by a circular area within the RNFL with posterior shadowing). The presence of this blood vessels has caused a localised thickening of the RNFL and is a normal variation.

RNFL thickness (heat) map

The RNFL heat map shows two areas of increased thickness superiorly rather than the one, single bundle typically noted.

RNFL deviation map

The RNFL deviation map flags an area of apparent thinning superiorly. This needs to be interpreted in conjunction with the heat map. The superior area is flagged as it is thinner that what would be expected within the age-matched normative database, however it is not a true thinning, rather a redistribution of the RNFL that has led to a thicker RNFL on either side.

TSNIT curve

The TSNIT curve shows a "double hump" superiorly (indicated by the blue arrows). This is referred to as a "split bundle" and is not indicative of true RNFL thinning but rather a physiological variation of the RNFL distribution. Some fibres that would normally be expected to be located superiorly have been shifted laterally. This has resulted in the superior RNFL being flagged as thin, however it is not representative of a true loss.

RNFL quadrant and clock hour analysis

The RNFL clock hour analysis has flagged superior thinning in both eyes with adjacent thickening due to the lateral redistribution of the RNFL.

RNFL thickness (heat) map

The heat map shows a temporal displacement of the RNFL thickness bundles.

RNFL deviation map

A temporal shift of the RNFL means that the areas superior and inferior to the disc are thinner than that expected age-matched normative thickness. As with the split bundle defect, this is not at true thinning, but a consequence of RNFL redistribution.

TSNIT curve

Note the temporal shift of the RNFL peak thicknesses both superiorly and inferiorly on the TSNIT curve.

RNFL quadrant and clock hour analysis

The RNFL clock-hour and quadrant analyses flag both superior and inferior RNFL thinning in each eye. However, the temporal shift of the RNFL bundles can be appreciated in the thicker than average temporal quadrant.

A typical healthy GCA thickness map (Right eye)

Note the typical "healthy" pattern of increased thickness around the macula (yellow) with thickness significantly reduced centrally (blue) due to the presence of the foveal pit.

GCA thickness map in a patient with superior arcuate VF defect

Note the inferior thinning (blue) while the superior GCIPL shows relatively normal thickness.

GCA thickness map in a patient with advanced glaucoma

Note the global thinning of the GCIPL.

A GCA deviation map showing no areas of loss (right eye)

Note the absence of yellow or red in this image indicating the GCIPL is within the age-matched normative range.

GCA deviation map in a patient with superior arcuate VF defect

Note the inferior thinning in an arcuate shape, sharply delineated at the horizontal raphe.

GCA deviation map in a patient with advanced glaucoma

Note the generalised reduction in GCIPL thickness

GCA global parameters in a healthy patient

Note that these parameters show symmetry between the two eyes and all metrics are flagged green indicating they are within normative limits.

GCA global parameters in a patient with superior arcuate VF defect

Note the average GCIPL thickness is still within the normative distribution. However, the minimum GCIPL thickness however is significantly reduced and flagged red.

GCA global parameters in a patient with advanced glaucoma (left eye)

Note the reduced average and minimum thicknesses in the left eye.

Sector analysis in a healthy patient

Note this analysis shows symmetry between the two eyes and all sectors are flagged green.

GCA sector analysis in a patient with superior arcuate VF defect

Note the thinning in the inferotemporal sectors.

GCA sector analysis in a patient with advanced glaucoma

Note the generalised reduction in thickness of the GCIPL

Fundus photograph and red-free image (left eye)

Retinal photos show a large optic nerve with thin superior neuroretinal rim with an adjacent RNFL defect (most evident on red-free imaging). There is also an apparent RNFL wedge defect inferiorly. .

Stereophoto (left optic disc)

Please use a stereoviewer or 15 prism dioptres of loose prism (base out) to view.

The stereophoto shows a large, deep cup and a focal loss of the superior neuroretinal rim. The lamina cribosa pores are visible and there is alpha peripapillary atrophy (PPA) present. There is no significant disc torsion or tilt.

Cirrus RNFL analysis

Signal strength is excellent and segmentation is accurate.

The heat map shows an asymmetry of the RNFL between the eyes with apparent thinning most notable superiorly in the left eye but also present inferiorly.

The TSNIT curve indicates that the RNFL thickness lies within the normative distribution in all aspects, however there is a relative supero-temporal thinning in the left eye compared with the right, and to a lesser degree, inferiorly also.

Examination of the raw data indicates a significant numerical difference between RNFL thickness in the right and left eye in the clock hours identified by the red squares in the image.

The thicknesses are still within the normative range, however given that they are significantly thinner than the corresponding clock hour thicknesses in the right, this is an indication of likely thinning.

Cirrus Ganglion Cell Analysis (GCA)

The GCA has flagged a small area of thinning infero-temporally in the left eye but is otherwise unremarkable.

Central 24-2 Threshold visual field

24-2 fields show an possible inferior depression in the left eye. With this case, while there is a demonstrable structural defect, function has not been significantly impacted.

Fundus photograph and red-free image (right eye)

Fundus photography shows a torted and tilted discs with large cups. The neuroretinal rim appears sloped superotemporally and inferotemporally but there is no focal notching or rim loss.

Cirrus RNFL analysis

RNFL heat and deviation maps show all parameters to be within the normative range in both eyes. However, as you can see, the individual RNFL thickness values are thinner in the superotemporal region in the right eye compared to the left eye (red squares) and, to a lesser extent, also thinner infero-temporally in the right eye compared with the left (purple squares).

Cirrus Ganglion Cell Analysis

The GCA shows the ganglion cell-inner plexiform layer (GCIPL) thickness in all sectors to be within the normative database, however comparison of the raw data shows a thinning supero-temporally in the right eye.

Central 24-2 threshold visual fields

Visual field results showed an inferior paracentral defect in the right eye which is concordant with the structural findings. Given these structural and functional findings, the right eye was suspicious for glaucoma. This is a case of green disease.

Fundus photographs (right and left eye)

Fundus photography shows tilted medium size discs in each eye with an intact neuroretinal rim. The RNFL shows a normal striated appearance with no areas of dropout evident.

Red-free images (right and left eye)

Red-free images shows a healthy RNFL reflectivity

RNFL analysis

The RNFL analysis flagged areas of superior thinning, most notable in the left eye. The TSNIT curve and clock hour analysis indicate that the superior RNFL thickness is outside the normative range, moreso the left than right eye.

Ganglion Cell Analysis

The GCA indicates a thinner than average GCIPL in both eyes, particularly inferiorly.

A common cause of thinner than average RNFL or GCIPL values is high myopia (Tai et al, 2018) and so RNFL and GCIPL thinning must be considered in the context of refractive status.

This patient has no other clinical indicators suspicious for glaucoma and no functional defects. Thus the RNFL and GCIPL findings are likely the result of the patient's high myopia. This is an example of red disease.

Fundus photographs (right and left eye)

Fundus photos show bilaterally small, crowded, tilted discs with small cups and a healthy neuroretinal rim in each eye.

Red-free images (right and left eye)

Red-free images show a normal, striated RNFL appearance.

Cirrus RNFL analysis

RNFL heat maps show a temporal displacement of the normal RNFL pattern in each eye. This displacement has caused the superior and inferior RNFL to be 'thinner than average' which is flagged on the deviation maps.

The TSNIT curve shows a symmetrical temporal shift of the superior and inferior RNFL in both eyes. This indicates that the 'thinning' flagged on the deviation maps is due to a temporal distribution of the RNFL and not indicative of true pathological loss.

Cirrus Ganglion Cell Analysis

GCA shows no significant thinning in either eye.

This is a young patient with no risk factors for glaucoma. Imaging confirms a temporal shift of the RNFL and there are no functional deficits. This is a case of red disease.