v1.0.1

Hemoglobin is the iron-containing protein, which carries oxygen in your bloodstream. RetinaLyze Glaucoma measures the relative hemoglobin amount in the optic disc in relation to the main retinal vessels. 

The greater the thickness of the tissue or its concentration of hemoglobin, the more intense the color, and the thinner the tissue or the less abundance of blood, the whiter it will appear. Thus, the concentration of hemoglobin in the optic disc can be used as an indication of the presence of glaucoma.

Heatmap

Below is an image, showing the difference in hemoglobin (relative to the vessels) between a normal optic disc and glaucomatous optic disc. 100% on the scale refers to the hemoglobin level in the central vessels of the retina. Warmer colors represent high densities of hemoglobin and cooler colors represent areas of lower perfusion or thin tissue.

Open

Normal (Left) and Glaucomatous (Right) optic nerve heads

RetinaLyze Glaucoma assesses the level of hemoglobin in the ONH (arteries, veins, rim and cup) by assessing the colours in the fundus image. The vessels are used as reference for calibration. 

Finally, a Globin Discriminant Function (GDF) is calculated to assess if there is damage to the ONH, which indicates the presence of glaucoma.

For a more in-depth explanation of the way the RetinaLyze Glaucoma algorithm works, check out the article below:

Read more about How the RetinaLyze Glaucoma algorithm works

Advanced Glaucoma Information

Advanced Glaucoma Information is a optional module, which can be enabled on a per account basis.

Open

Globin Discriminant Function (GDF)

The most validated index is the GDF. The value 0 has been adjusted to an approximate specificity of 95%. More positive values indicate more normal perfusion, and negative values indicate poorer or less frequent perfusion.

This metric is especially useful in glaucoma screening, where the 1% and 2% percentiles for normal individuals are -15 and -10, respectively.

Vertical Cup/Disc Ratio & Cup/Disc Area Ratio

These ratios are derived from analyzing the optic cup's size, shape, and position through the distribution of hemoglobin within the eye's image. By comparing these ratios with those from a general population, we can identify atypical structures, assisting in glaucoma diagnosis. The 1% and 2% percentile benchmarks are provided for context.

Glaucoma usually preferentially affects the upper and lower poles of the nerve, which generally leads to vertical cup growth.¹

Advanced Overlays

Cup and disc segmentation

Shows the original retinal photo with automatic segmentation of the optic disc, approximately at the border of Elschnig's scleral ring. It also shows an estimation of the position of the cup, based on the hemoglobin distribution.

Hemoglobin pseudo-color map

Shows vessel segmentation, the color of which represents 100% hemoglobin, and the percentage estimate of hemoglobin in the disc tissue.

Sectors area according to Hb

Shows, in percentage, the area of the cup and of each sector of the optic disc relative to the total area of the disc (100%). A chromatic code shows the patient's area in relation to the usual percentiles in the normal population. The sectors of the rim correspond to the following angular positions.

Globin Individual Pointer (GIP)

GIP is a optional module, which can be enabled on a per account basis. It’s primarily aimed at ophthalmologists.

GIP (Glaucoma Index of Progression) is an analytical tool designed to measure changes in glaucoma status with notable stability, albeit with slightly less precision than GDF (Globin Discriminant Function) in determining the exact boundary of normalcy. This difference primarily arises from the reduced influence of deep learning algorithms in GIP's calculation, making GIP particularly useful for tracking the progression or regression of a condition over time. It's crucial to remember that "normality" is statistically defined, and a patient's condition can worsen even if their measurements remain within normal ranges.

For instance, consider height as an analogy: A height of 1.70 meters might be within the normal range, but if the individual was 1.80 meters tall a year earlier, this change suggests an underlying issue, despite the current height still being "normal."

Image Saturation

This refers to the intensity of color and light in the captured image. The system can process images with varying degrees of saturation and provide results, but we recommend using lower flash levels for optimal image quality. This adjustment ensures that the diagnostic tools can analyze the images more effectively.

Image Quality

The system evaluates the quality of each image before analysis. Poor-quality images, especially those where the optic disc is partially or completely obscured, are excluded from analysis to maintain accuracy in diagnostics. If the image is of sufficient quality, the system informs the user of its evaluative judgment, ensuring that only reliable data is used for assessment.

Disc Area

The analysis of the optic disc area is performed by comparing the test results against a database of tests conducted with the same type of camera. This comparison is expressed in mm2, providing a benchmark for evaluating the optic disc's size relative to a normative database. This metric is valuable for identifying deviations from typical disc sizes, which can be indicative of glaucoma or other optic nerve issues.

Disc size plays a role in the evaluation of these ratios. Especially very large optic nerve heads tend to have large cups, often difficult to differentiate from glaucomatous cups. Approximately 5-6% of the nerves are larger than 2.5mm2. The results should be interpreted with caution in these cases.

RetinaLyze Glaucoma - Perimetry combined analysis

Combining GDF with Perimetric Indices The Globin Discriminant Function (GDF) can be enhanced by incorporating perimetric data, offering a more comprehensive view of the patient's ocular health. To integrate this data, select the "Add VF indexes" option. This allows for a multifaceted analysis by combining GDF with visual field (VF) indices.

Understanding Mean Defect (MD) and Its Application The Mean Defect (MD) is a critical parameter used to adjust for the bi-linear response observed in PSD-sLV (Pattern Standard Deviation-standardized Loss Variance) calculations. When interpreting results from Octopus perimeters, it's important to invert the MD values, as their scale is opposite compared to other perimeter types. Essentially, MD helps to linearize the data for consistent analysis across different testing devices.

Introducing the TCV Index for Perimetric Analysis As an enhancement to PSD-sLV, the Thresholds Coefficient of Variation (TCV) offers an alternative method focusing on perimetric harmony. TCV assesses variability using data from 18 specific points in the visual field. To utilize TCV, manual input of these 18 threshold values, as indicated in a provided image, is required. This index allows for a targeted examination of visual field consistency, aiding in the detection and monitoring of glaucoma progression.:

This diagnostic approach has been specifically tested with two types of visual field testing equipment and methodologies:

• Octopus Perimeters using the Tendency Oriented Perimetry (TOP) Strategy: This is a rapid visual field testing method that takes advantage of the trends or relationships between the sensitivities of the points in the visual field, conditioned by the axon paths of the ganglion cells, to make a rapid assessment of individual visual field (not its progression).

• Oculus Perimeters with Spark Strategy: Another quick testing method that uses a distinct approach to analyze the visual field, suitable for detecting early changes.

Incorporating Humphrey Visual Field Data Data from Humphrey perimeters, which use the SITA (Swedish Interactive Threshold Algorithm) strategy, can also be integrated into this method. Although SITA and TOP are different, our process adapts Humphrey's data to make it compatible with TOP-based analyses. This allows for a broader application of the method across different types of perimetry data.

User Evaluation of RetinaLyze Glaucoma Integration The effectiveness of combining this method with RetinaLyze Glaucoma, a software for analyzing retinal images for signs of glaucoma, should be assessed by clinicians themselves. This encourages users to critically evaluate how the integration of visual field data with retinal imaging can enhance glaucoma detection and monitoring in their practice.

Timeline

Simultaneous Evaluation of Bilateral Eye Changes This feature enables the comparison of changes in the visual appearance of both eyes over time by analyzing and displaying the results side by side. It focuses on examining the GIP (Glaucoma Index of Progression) indices along with the linearly estimated areas of the optic cup and rim sectors. This allows for a comprehensive overview of the progression or regression of conditions affecting the optic nerve and its surrounding structures.

Age Independence in Analysis Prior research has indicated that these changes are not significantly affected by the patient's age. Consequently, the analysis utilizes linear regression based on the date of the first available image of the patient. This approach ensures that the evaluation is temporal, focusing on changes over time rather than age-related differences.

Interpreting the Regression Analysis The results include the formula for the GIP regression line, which calculates the rate of change over time, and its statistical significance, indicated by the P value. The P value helps determine whether the observed changes are statistically meaningful. It's important to note that the reliability of these results is influenced by both the extent of change observed in the patient's condition and the number of examinations analyzed. While a minimum of three examinations is required to perform this analysis, a higher number of exams generally provides a more robust basis for statistical significance.

The XY plot visually represents the regression line that illustrates the progression or regression of GIP (Glaucoma Index of Progression) values over time for a patient. The plot uses a color-coding system to indicate how the patient's GIP values compare to those in a normal reference population:

• Green: Indicates GIP values greater than the 5th percentile, suggesting the patient's measurements are within normal limits for the majority of the population.

• Yellow: Represents GIP values that fall between the 5th and 1st percentiles, indicating a borderline or marginal condition that warrants closer observation.

• Red: Signifies GIP values below the 1st percentile, highlighting a significant deviation from normal values and suggesting a higher risk or presence of disease.

This color-coding helps in quickly identifying the patient's standing in relation to typical GIP values and assists in determining the urgency and type of intervention needed.

The displayed middle images provide a visual summary of how the areas of the optic cup and rim sectors have changed at each evaluated point in time. These images are designed to offer a quick overview of the progression or regression of these critical areas, which are vital for assessing glaucoma and other optic nerve health issues.

The images located at the bottom part of the display are designed to provide a visual representation of the changes in the areas of the optic cup and the rim over time. These changes are crucial for monitoring conditions such as glaucoma, where alterations in these structures can indicate disease progression.

Each image employs a color-coding system to highlight the statistical significance of the observed changes, as determined by the regression analysis for both the cup and the rim areas. This chromatic code enables quick identification of how significant these changes are, according to standard statistical measures. Essentially, the colors in these images help in assessing whether the changes over time are likely due to natural variations or are indicative of a more significant trend that may require clinical attention.

Sources

1. ‘10 | Glaucoma 2022–2023 BCSC Basic and Clinical Science Course’ - Page 64, 116 and 120 by American Academy of Ophthalmology (Last major revision 2020–2021)