Who should use the Automated Eye Disease Screening workflow?
Teams or solo builders working on healthcare tasks who want a repeatable process instead of one-off tool experiments.
AI Workflow · Healthcare
Leverage AI-powered analysis of retinal images and OCT scans to detect diabetic retinopathy, glaucoma, and age-related macular degeneration, with automated triage and referral for timely care.
Deliverable outcome
A clinician-validated report and referral order in the EHR.
30-90 minutes
Includes setup plus initial result generation
Free to start
You can swap tools by pricing and policy requirements
A clinician-validated report and referral order in the EHR.
Use each step output as the input for the next stage
Step map
Instead of relying on a single generic AI model, this pipeline connects specialized tools to maximize quality. First, you'll use EyeArt AI Eye Screening System to a validated set of high-quality retinal and oct images ready for analysis. Then, you pass the output to Cellpose to pixel-level masks of all relevant anatomical and pathological features. Then, you pass the output to RetiSpec to a structured set of biomarker values with severity scores for each disease. Then, you pass the output to EyeArt AI Eye Screening System to a per-disease severity grade (e.g., moderate dr, suspect glaucoma, early amd). Then, you pass the output to Digital Diagnostics (formerly IDx) to a triage level and a draft referral note for each patient. Finally, Digital Diagnostics (formerly IDx) is used to a clinician-validated report and referral order in the ehr.
Acquire and Validate Retinal Images and OCT Scans
A validated set of high-quality retinal and OCT images ready for analysis.
Segment Anatomical Structures and Lesions
Pixel-level masks of all relevant anatomical and pathological features.
Extract Disease-Specific Biomarkers
A structured set of biomarker values with severity scores for each disease.
Classify Disease Presence and Severity
A per-disease severity grade (e.g., moderate DR, suspect glaucoma, early AMD).
Generate Triage and Referral Recommendations
A triage level and a draft referral note for each patient.
Review and Finalize Report (Optional Human-in-the-Loop)
A clinician-validated report and referral order in the EHR.
Collect high-resolution retinal fundus photographs and optical coherence tomography (OCT) scans from the patient. Ensure each image meets quality standards (e.g., sufficient illumination, no artifacts, correct field of view) by running automated quality assessment algorithms. Reject or flag poor-quality images for retake before proceeding.
Why EyeArt AI Eye Screening System: EyeArt AI Eye Screening System is designed for retinal image analysis and can detect diabetic retinopathy, macular degeneration, and glaucoma, making it suitable for acquiring and validating retinal images and OCT scans.
Use a segmentation neural network (e.g., U-Net or transformer-based) to delineate key retinal landmarks: optic disc, macula, blood vessels, and fluid/lesion regions (exudates, hemorrhages, drusen). For OCT, segment retinal layers and identify fluid pockets. Output pixel-level masks for each structure.
Why Cellpose: Cellpose is a general-purpose segmentation AI model that can segment anatomical structures and lesions in medical images, including retinal scans, and supports 3D volumetric segmentation.
From the segmented masks, compute quantitative biomarkers: cup-to-disc ratio (glaucoma), number and area of microaneurysms/hemorrhages (diabetic retinopathy), drusen volume and geographic atrophy area (AMD). Also extract OCT-derived features like central subfield thickness and fluid volume. Normalize values against patient demographics.
Why RetiSpec: RetiSpec specializes in extracting Alzheimer's disease biomarkers from retinal imaging using hyperspectral analysis, directly matching the need for biomarker extraction from retinal scans.
Feed the extracted biomarkers and raw image features into a multi-label classification model (e.g., ensemble of CNNs and transformer). The model outputs probabilities for diabetic retinopathy (none/mild/moderate/severe/proliferative), glaucoma (suspect/moderate/advanced), and AMD (early/intermediate/late). Apply a confidence threshold to flag uncertain cases for manual review.
Why EyeArt AI Eye Screening System: EyeArt AI Eye Screening System provides multi-label classification for diabetic retinopathy, macular degeneration, and glaucoma, with severity grading capabilities.
Combine severity grades with patient history (e.g., diabetes duration, intraocular pressure, family history) using a rule-based or learned triage system. Output one of: routine follow-up, urgent referral (within 1 week), or emergency referral (within 24 hours). Include specific recommendations (e.g., 'refer to retina specialist for anti-VEGF therapy').
Why Digital Diagnostics (formerly IDx): Digital Diagnostics offers autonomous clinical decision support, which includes generating triage and referral recommendations based on retinal screening results.
Present the AI-generated findings, biomarkers, and triage recommendation to a clinician (ophthalmologist or optometrist) via a dashboard. The clinician can accept, modify, or override the AI output. Once approved, the final report is signed and stored in the electronic health record (EHR).
Why Digital Diagnostics (formerly IDx): Digital Diagnostics provides autonomous clinical decision support with a clinician review interface and can integrate with EHR systems for report finalization.
§ Before you start
Teams or solo builders working on healthcare tasks who want a repeatable process instead of one-off tool experiments.
No. Start with the top pick for each step, then replace tools only if they do not fit your pricing, compliance, or output needs.
Open the mapped task page and compare top options side by side. Prioritize output quality, integration fit, and predictable cost before scaling.
§ Related
Track competitor moves and market shifts in real-time with automated intelligence gathering — so you always know what your rivals are doing.
Connect siloed business applications into a unified, AI-managed operational pipeline that eliminates manual handoffs between systems.
Analyze portfolios, backtest investment strategies, and receive AI-generated market signals — giving individual investors access to institutional-grade tools.