EQA Grading Guidance

About this platform

EQA — External Quality Assurance Platform is a medical educational workspace for diabetic retinopathy grading, retinal image review, consensus development, and quality assurance reporting. It is designed to support structured learning, repeatable grading, and clinically meaningful feedback across Level 1 and Level 3 workflows.

1. Review the image carefully before grading

Always examine the retinal image first before selecting any grade. Use the available image tools to inspect subtle lesions, especially when image quality is reduced or lesions sit at the edge of the field.

2. Use the image tools

The EQA grading screen includes image controls for:

• Zoom in / zoom out / reset for closer inspection
• Black & white mode
• Negative mode
• Contrast enhancement for difficult images

These tools are intended to help reveal microaneurysms, haemorrhages, exudates, vessel changes, and subtle boundary detail. Be especially cautious near the edge of the photographed field.

3. Adjust contrast and brightness thoughtfully

If an image looks hazy or low contrast, increasing contrast is often more useful than simply increasing brightness. In many clinically difficult images, the best view comes from a combination of slightly reduced brightness and increased contrast.

4. Use the fovea / disc measurement tool

The platform includes a Fovea tool for placing the fovea, disc edge A, and disc edge B. This helps estimate disc-based distance relationships on the image.

As in clinical grading practice, distance judgements matter when deciding whether lesions lie within one or two disc diameters of the fovea.

5. Use the mask tool for training annotations

The Mask tool can be used to draw labelled annotation boxes over lesions or structures of interest. These saved annotations can later be reviewed visually, which is useful for education, consensus review, and auditing of grading decisions.

6. Quality grading and technical failure

Image quality should be judged carefully. However, when clearly referable pathology is visible, that pathology should still be graded even if image quality is otherwise poor. In practice, visible referable disease still matters clinically and educationally.

Assess quality in terms of both:

• field definition — was the correct retinal area captured?
• clarity — would important pathology be visible if present?

7. Dots, blots, and microaneurysms

Small red lesions require careful interpretation. In general:

• Microaneurysms are usually small, round, and well defined
• Blot haemorrhages are typically larger and less sharply defined

When uncertain, use the image tools, zoom, and contrast adjustments before grading.

8. Save and navigation behaviour

In this EQA build, answers are saved when you move to the next question or when the workflow performs its submission/save step. If you remain in the assignment, you can resume where you left off. Review your selections before final submission.

9. Referral logic and consensus

This platform supports structured referral logic and gold-standard comparison. In the current EQA setup, referable status may be driven by the configured retinal / maculopathy / quality thresholds for a given test. Review pages may show:

• referable vs non-referable outcome
• question-level agreement
• field-level agreement
• ROC-oriented performance views

10. Practical advice

• Inspect the image before grading
• Use zoom and contrast tools actively
• Be cautious at the edge of the field
• Use the fovea/disc tool when distance matters
• Use masks to highlight training-relevant lesions
• Grade visible disease even when quality is imperfect
• Use review pages to learn from agreement, outcome, and consensus feedback