Skip to content

How RetGuard Works

From image capture to clinical report in under 30 seconds. Follow the three steps of a RetGuard screening.

Step 1 — Capture

Two imaging modalities. One device.

The RetGuard camera captures both fundus photographs and OCT scans — the two gold-standard imaging methods for retinal screening — giving the AI a complete picture of the retina.

Fundus photograph showing retinal surface with blood vessels, optic disc, and macula
Fundus Photography
Surface view of blood vessels, optic disc, and macula
DRGlaucomaHTR
OCT cross-sectional scan showing retinal layers and foveal structure
Optical Coherence Tomography
Cross-sectional view of retinal layers and structural changes
DMEAMD
Built-in quality checks verify image clarity and coverage before any analysis begins.
Step 2 — Analyze

Five conditions. One simultaneous analysis.

RetGuard's AI analyzes the captured images and screens for all five conditions at once — in under 30 seconds, entirely on-device, with no data leaving the clinic.

DRFundus

Diabetic Retinopathy

Microaneurysms and dot hemorrhages
Hard and soft exudates
Neovascularization and vessel changes
DMEOCT

Diabetic Macular Edema

Macular thickening and fluid accumulation
Cystoid spaces in retinal layers
Subretinal fluid near the fovea
GLCFundus

Glaucoma

Increased cup-to-disc ratio
Retinal nerve fiber layer thinning
Optic disc rim loss and notching
AMDOCT

Age-Related Macular Degeneration

Drusen deposits in the macula
Pigment epithelium changes
Choroidal neovascularization signs
HTRFundus

Hypertensive Retinopathy

Arteriolar narrowing and AV nicking
Flame-shaped hemorrhages
Cotton wool spots and papilledema

All five conditions are screened from a single capture session. Results include calibrated confidence scores and referral recommendations for each condition independently.

Step 3 — Report

Every result comes with visual evidence

RetGuard generates visual heatmaps that highlight the exact retinal regions that influenced each screening result — so clinicians can see what the AI found and where.

When Hypertensive Retinopathy is detected, the report automatically recommends relevant laboratory tests — bridging retinal findings with systemic evaluation.

Per-condition heatmap overlays on retinal images
Calibrated confidence scores with referral recommendations
Automated lab test recommendations when HTR is detected
Structured PDF report ready for clinical records
RetGuard Screening Report
Sample output
Diabetic Retinopathy
97%Refer
Glaucoma
12%Normal
Diabetic Macular Edema
84%Refer
AMD
8%Normal
Hypertensive Retinopathy
68%Refer
Laboratory Tests Recommended
HTR detected — recommend blood pressure measurement, lipid panel, HbA1c, and renal function tests to assess systemic hypertensive status.
Grad-CAM Evidence Attached
3 heatmaps · DR, DME, HTR regions highlighted