Oku Digital Health, a La Jolla, Calif., company spun out of Topcon Healthcare, introduced Connected Care on Oct. 6, a service that captures retinal images during routine eye exams and returns AI-supported insights on heart and other systemic disease risk. The company says it reaches more than 10,000 care locations, but its launch materials do not claim FDA clearance for an Oku-branded cardiovascular risk model.
That gap shapes how the product should be read, and it sits beside a broader claim: that the eye can serve as an entry point for catching disease well beyond ophthalmology.
How Connected Care works
According to Oku's announcement, the software captures retinal images at the point of care during primary or comprehensive eye visits. It then delivers AI-supported insights into the eye care practice's electronic health record workflow, and higher-risk patients are routed on to appropriate care. It integrates with Topcon Harmony, which combines imaging with clinical data, and offers AI-enabled triage and decision dashboards.
Oku said the analysis looks for signs associated with cardiovascular disease, chronic kidney disease, hypertension, diabetic retinopathy, glaucoma and age-related macular degeneration. Biological age insights are also part of the offering.
Image ingestion works with equipment from any vendor, the company said, though it is most tightly integrated with Topcon devices. The product supports two-way referral and co-management with specialists. Oku also said it supports HEDIS EED and Medicare Advantage Star measures, the quality benchmarks health plans use to judge things like diabetic eye exam completion.
Ali Tafreshi, Oku's CEO, framed the pitch around both sides of the referral relationship. "Primary care practices face a significant operational burden in closing quality care gaps, while eye care practices are looking for ways to expand their clinical role and strengthen patient relationships," he said.
What the company does not claim
Oku's materials say the AI looks for "signs associated with cardiovascular disease." They stop short of claiming FDA clearance for cardiovascular risk prediction. A provider FAQ says autonomous AI diabetic retinopathy results are available within Oku workflows, yet the company does not present an FDA-cleared, Oku-branded cardiovascular model.
The nearest related asset belongs to Toku. Topcon signed a definitive agreement to acquire that company on Oct. 17, 2025. Toku's CLAiR AI uses retinal images to assess cardiovascular risk, biological age and chronic kidney disease, and the company obtained CE and UKCA marks, the European and British conformity approvals, for AI cardiovascular risk assessment on Feb. 5, 2024. I found no announcement of U.S. FDA clearance for Toku's cardiovascular product as of Oct. 31, 2026.
Reach, and a numbers discrepancy
Oku said it draws on Topcon's global installed base of more than 50,000 diagnostic devices, and that more than 2.2 million patients outside traditional eye care have been screened to date. Its homepage tells a slightly different story, citing 2.5 million or more patients screened, attributed to "company data on file" from June 2026. The company has not explained the difference.
The homepage also lists more than 10,000 provider locations, split between more than 8,000 primary eye care sites and more than 2,000 primary care sites.
A parallel product for drug developers
The same day, Oku launched a second offering aimed at pharmaceutical sponsors. It includes AI-assisted reading center services, point-of-care patient identification for clinical trials, and site selection based on verified patient density.
The company said its data assets include 17.1 million retinal images and 1.2 million OCT volumes (optical coherence tomography, a layered scan of the retina). Tokenized longitudinal EHR linkages span 14 health systems and 146 million exams, it said.
"Pharma sponsors face enormous pressure to identify the right patients faster and reduce screen-fails across fragmented trial workflows," said Trent Norris, Oku's chief operating officer.
The platform runs on Microsoft Azure and is optimized with NVIDIA CUDA libraries. Oku said it uses tenant isolation and end-to-end encryption, has HITRUST and ISO 27001-certified components, and links de-identified imaging to EHR data through tokenization.
Leadership and ownership
Tafreshi also serves as CEO of Topcon Healthcare. Norris previously worked at NVIDIA, Flywheel.io and Microsoft. Matthew Lungren, MD, MPH, is chief data and AI officer; his Stanford profile lists him as an adjunct and a former Microsoft Health & Life Sciences executive. Daniela Ferrara, MD, PhD, is chief medical officer.
The founding team also includes Jacques Gilbert, chief strategy and business development officer; Ray Everett, general counsel and chief data protection officer; and Tomomichi Sassa, chief financial officer.
Oku has disclosed no independent venture financing as of Oct. 31, 2026. Launch materials call it a "separate company," but its equity structure and capitalization have not been detailed, which leaves open how independent it is from its Topcon parent in practice.
A crowded field
Oku is not alone in chasing the retina's systemic signal. Eyenuk's materials describe autonomous diabetic retinopathy screening, with a roadmap toward predictive biomarkers for systemic cardiovascular and neurological health. AEYE Health published a March 2026 release on non-invasive blood pressure estimation using retinal AI; the product was not FDA-cleared for blood pressure or cardiovascular use at that time.
Clinical evidence is also accumulating. A March 30, 2026 AHA/ACC news release described a prospective multicenter trial of AI retinal imaging for elevated ASCVD risk, the kind of heart disease driven by arterial plaque buildup. It said CLAiR flags a 10-year risk threshold of 7.5% or higher, with about five minutes of imaging and roughly 30 seconds of inference.
Oku made its public debut at the American Academy of Ophthalmology meeting in New Orleans, Oct. 9-12, at Topcon booth #1737.
Tafreshi, speaking to Women in Optometry, put the ambition plainly. "The eye is becoming the front door to medicine, but a front door only matters if it leads somewhere," he said.
