AI optical coherence tomography screening

01Overview and value chain

Markers: [EC: IVDR 2017/746 + MDR 2017/745 | OECD: Biotech-health | Regulator: FDA (USA), EMA (European Union), NMPA (China)]

AI optical coherence tomography (OCT) screening applies machine-learning algorithms to OCT and fundus imaging for broad eye-disease detection — diabetic retinopathy, age-related macular degeneration (AMD) and glaucoma — distinct from the neurodegenerative-brain-disease-specific retinal biomarker research covered elsewhere in this catalog. Notal Vision’s Home OCT system lets AMD patients self-monitor at home, with a pivotal study validating diagnostic equivalence to in-clinic OCT and reporting that the resulting data measurably influences physician treatment decisions. Topcon Healthcare’s Maestro2 multimodal platform combines OCT and color fundus imaging, and recently received FDA 510(k) clearance for robotic OCT angiography, adding AI-assisted vascular imaging to its screening capability. RetinAI, a Swiss AI-based OCT analytics platform, was acquired by EssilorLuxottica in a deal reported in late 2025, now operating as part of the eyewear and eye-health giant’s broader portfolio. In China, Airdoc develops AI-based eye-screening technology as part of the country’s growing ophthalmic AI diagnostics base. In the US, iHealthScreen recently received FDA 510(k) clearance (K253704) for iPredict DR, AI-powered software for automated diabetic retinopathy screening.

The key directions of AI optical coherence tomography screening are:

1. Home/remote OCT monitoring (Home OCT Monitoring): patient-operated OCT devices enabling self-monitoring outside the clinic, validated against in-clinic diagnostic accuracy — Notal Vision.
2. Multimodal AI OCT platforms (Multimodal AI OCT Platforms): clinic-based systems combining OCT, fundus photography and increasingly OCT angiography with AI-assisted interpretation — Topcon Healthcare’s Maestro2.
3. AI OCT analytics software (AI OCT Analytics Software): software platforms applying machine learning to OCT image analysis, now often integrated into larger eye-health or eyewear businesses — RetinAI.
4. Autonomous AI diabetic retinopathy screening (Autonomous AI DR Screening): FDA-cleared AI software that independently interprets fundus/OCT images to flag diabetic retinopathy without requiring a specialist to read every image — iHealthScreen, Airdoc.

Sectoral value chain

[OCT/fundus image capture] ──> [AI algorithm disease detection] ──> [screening result/report generation]
                                                    │
                                       (regulatory clearance: FDA/MDR-IVDR/NMPA)
                                                    │
                                                    ▼
[population screening program integration] <─── [clinical referral decision] <─────┘

Value chain levels

Cross-cutting technologies of the sector:

• Home-use OCT hardware (Home OCT Monitoring): simplified, patient-operable OCT devices extend monitoring frequency beyond periodic in-clinic visits, validated to match in-clinic diagnostic accuracy.
• Multimodal imaging integration (Multimodal AI OCT Platforms): combining OCT, fundus photography and OCT angiography in one platform, with AI assisting interpretation across modalities.
• Autonomous AI diagnostic software (Autonomous AI DR Screening): FDA-cleared algorithms that independently classify disease presence from an image, reducing reliance on specialist image review for population-scale screening.

02US

The US anchors the deepest bench of AI-OCT screening companies, spanning home monitoring devices and autonomous diagnostic software.

Notal Vision, iHealthScreen, FDA

• Notal Vision: its Home OCT system lets AMD patients self-monitor at home, with a pivotal study validating diagnostic equivalence to in-clinic OCT and reporting that the resulting data measurably influences physician treatment decisions.
• iHealthScreen: recently received FDA 510(k) clearance (K253704) for iPredict DR, AI-powered software for automated diabetic retinopathy screening.
• FDA framework: AI-OCT screening devices and software clear the FDA via 510(k) or De Novo review, with autonomous diagnostic claims (iPredict DR) requiring specific clinical validation against specialist-graded reference standards.

03CN

China develops AI-based eye-screening technology through domestic providers including Airdoc, part of the country’s growing ophthalmic AI diagnostics base.

Airdoc, NMPA

• Airdoc: develops AI-based eye-screening technology, including diabetic-retinopathy and broader ophthalmic disease detection, as part of China’s growing ophthalmic AI diagnostics base.
• NMPA framework: AI-assisted OCT and fundus-screening software in China requires NMPA approval as a medical device, with the AI algorithm component often reviewed alongside the imaging hardware.

04EU

Switzerland anchored a European AI-OCT analytics entrant through RetinAI, though the company was acquired by EssilorLuxottica in a deal reported in late 2025, now operating within the eyewear and eye-health group’s broader portfolio.

RetinAI, EssilorLuxottica, MDR/IVDR

• RetinAI: its AI-based OCT analytics platform (RetinAI Discovery) was acquired by EssilorLuxottica, a transaction reported in October 2025, positioning the technology within a much larger eye-health and eyewear business.
• MDR/IVDR framework: AI-OCT screening devices and diagnostic software sold in the EU require conformity assessment under the Medical Device Regulation (MDR 2017/745), with any paired in-vitro diagnostic software component additionally subject to IVDR 2017/746.

05Leading companies and research institutes

06Tech stack and innovations

The stack spans home-use imaging hardware, multimodal clinic platforms and autonomous AI interpretation software, each extending OCT/fundus screening capacity beyond what specialist-only image review could deliver at population scale.

1. Home-use OCT devices (Home OCT Monitoring):
   • Notal Vision’s Home OCT system, validated for diagnostic equivalence to in-clinic OCT, lets AMD patients generate clinically actionable imaging data between office visits.
2. Multimodal clinic platforms (Multimodal AI OCT Platforms):
   • Topcon Healthcare’s Maestro2 combines OCT, color fundus photography and now FDA-cleared robotic OCT angiography in one system, with AI assisting interpretation across the combined imaging modalities.
3. AI OCT analytics software (AI OCT Analytics Software):
   • RetinAI Discovery applies machine learning to OCT image analysis, technology EssilorLuxottica’s 2025 acquisition folded into a global eye-health and eyewear portfolio.
4. Autonomous AI diagnostic screening (Autonomous AI DR Screening):
   • iHealthScreen’s iPredict DR (FDA-cleared) and Airdoc’s AI eye-screening technology both independently classify diabetic retinopathy presence from imaging, reducing the specialist-review bottleneck for population-scale screening.

07Value chains and production pipelines

Industrial pipeline of an AI-OCT screening system (FDA 510(k)/De Novo / EU MDR-IVDR / NMPA)

┌───────────────────────────┐      ┌───────────────────────────┐
│ 1. Image capture           │ ───> │ 2. AI disease detection  │
└───────────────────────────┘      └───────────────────────────┘
                                                   │
                                                   ▼
┌───────────────────────────┐      ┌───────────────────────────┐
│ 4. Regulatory clearance   │ <─── │ 3. Screening result/     │
│                            │      │    report                   │
└───────────────────────────┘      └───────────────────────────┘
              │
              ▼
┌───────────────────────────┐      ┌───────────────────────────┐
│ 5. Clinical referral      │ ───> │ 6. Population screening  │
│    decision                 │      │    integration               │
└───────────────────────────┘      └───────────────────────────┘

Stage 1: Image capture

OCT and/or fundus imaging is performed either in a clinic (Topcon Healthcare’s Maestro2) or by the patient at home (Notal Vision’s Home OCT), capturing the retinal or macular structure needed for AI analysis.

Stage 2: AI disease detection

A machine-learning algorithm analyzes the captured image for signs of diabetic retinopathy, AMD or glaucoma, the step where iHealthScreen’s iPredict DR and Airdoc’s screening technology operate autonomously.

Stage 3: Screening result and report

The AI classification is translated into a screening result or clinical report, formatted for either specialist review or, for autonomous systems, direct patient/primary-care communication.

Stage 4: Regulatory clearance

The imaging device and its AI software clear FDA 510(k)/De Novo review, EU MDR/IVDR conformity assessment, or NMPA registration, with autonomous diagnostic claims requiring validation against specialist-graded reference standards.

Stage 5: Clinical referral decision

A positive screening result triggers referral to an ophthalmologist for confirmatory examination and treatment planning, while a negative result returns the patient to routine screening intervals.

Stage 6: Population screening integration

Health systems deploy the cleared technology into routine screening programs for diabetic or aging populations, extending specialist-equivalent screening capacity to settings that would otherwise lack ophthalmologist access.