A single photograph of the fundus reveals diabetic retinopathy, glaucoma, macular degeneration, hypertensive disease, and cardiovascular risk — all in a noninvasive, 90-second exam. Sentinel Visio reads every pixel with the precision of a retina specialist and the speed of an algorithm.
Diabetic retinopathy is the leading cause of blindness in working-age adults — and nearly all of it is preventable with early detection. Yet 50% of the 537 million people with diabetes worldwide do not receive annual retinal screening. The barrier is not technology — it is access. There are not enough ophthalmologists to screen every diabetic patient, especially in primary care settings, rural communities, and developing nations. Autonomous AI eliminates this bottleneck entirely. A non-mydriatic fundus camera and Sentinel Visio can screen a patient in 90 seconds — without an ophthalmologist, without pupil dilation, without a referral — and deliver an immediate, clinically actionable result.
But Sentinel Visio goes far beyond diabetic retinopathy. The retina is a window into the entire body. Its vasculature reflects systemic hypertension, cardiovascular disease, and cerebrovascular risk. Its nerve fiber layer reveals glaucoma. Its macula shows age-related degeneration. A single fundus photograph, read by AI, becomes a multi-disease screening platform that transforms a routine diabetes visit into a comprehensive vascular health assessment.
Sentinel Visio detects and grades every stage of diabetic retinopathy — enabling intervention at the optimal moment for each patient.
Sentinel Visio transforms the fundus photograph from a single-disease screening tool into a comprehensive vascular health assessment.
This is the core engine — the one that eliminates the screening bottleneck. Sentinel Visio autonomously analyzes fundus photographs from non-mydriatic cameras, detecting microaneurysms, hemorrhages, hard exudates, cotton-wool spots, venous beading, IRMA, and neovascularization with sensitivity and specificity that parallels expert ophthalmologist grading. The system classifies DR into five ICDR stages (no DR, mild NPDR, moderate NPDR, severe NPDR, proliferative DR) and detects diabetic macular edema. Positive findings trigger automated referral pathways. Negative findings confirm safe rescreening intervals — all without an ophthalmologist touching the case.
Diabetic macular edema is the most common cause of visual impairment in patients with DR — and it can occur at any stage of retinopathy, including mild NPDR. DME causes central vision loss through fluid accumulation in the macula, and it responds well to anti-VEGF therapy when detected early. Sentinel Visio detects DME from both fundus photography (hard exudate patterns, macular thickening signs) and, when available, OCT integration (quantitative retinal thickness measurements, subretinal fluid, cystoid spaces). The system flags center-involving DME for urgent ophthalmology referral while monitoring non-center-involving DME longitudinally.
Glaucoma is the leading cause of irreversible blindness worldwide — and 50% of people with glaucoma are undiagnosed because the disease is asymptomatic until significant, irreversible nerve damage has occurred. Since every fundus photograph Sentinel Visio captures for DR screening also images the optic disc, the system simultaneously evaluates glaucoma risk: analyzing cup-to-disc ratio, neuroretinal rim thinning, disc hemorrhages, and peripapillary atrophy. Patients flagged as glaucoma-suspect receive automated referral for IOP measurement and visual field testing — transforming a DR screening visit into a dual-disease detection opportunity.
Every fundus photograph captured by Sentinel Visio also reveals the macular architecture, enabling simultaneous screening for age-related macular degeneration. The system detects drusen (the hallmark early finding), pigmentary changes, geographic atrophy (dry AMD), and signs of neovascularization (wet AMD) — stratifying patients into risk categories and recommending appropriate follow-up intervals, nutritional supplementation (AREDS2 formula), and urgent referral for suspected wet AMD requiring anti-VEGF therapy.
The retinal vasculature is the only vasculature in the human body that can be directly visualized without invasive procedures. Sentinel Visio's vascular analysis engine measures arteriolar and venular caliber, arteriovenous ratio, vessel tortuosity, branching patterns, and wall-to-lumen ratio — biomarkers that independently predict hypertensive end-organ damage, cardiovascular event risk, stroke risk, chronic kidney disease progression, and cognitive decline. A fundus photograph becomes a vascular health assessment that extends far beyond the eye.
Anti-VEGF therapy has revolutionized the treatment of diabetic macular edema, wet AMD, and retinal vein occlusion — but the injection burden is enormous: monthly or bimonthly injections, often for years. Sentinel Visio analyzes serial OCT and fundus imaging to predict treatment response, identify early non-responders who may benefit from switching agents, and optimize injection intervals through treat-and-extend protocols. The system monitors for recurrence patterns, enabling proactive retreatment before vision loss recurs rather than reactive retreatment after it has already happened.
The greatest impact of AI in ophthalmology is not in the retina specialist's office — it is in the primary care clinic, the endocrinology practice, the community health center, and the rural pharmacy where diabetic patients actually receive their care. Sentinel Visio is designed for point-of-care deployment: compatible with portable non-mydriatic fundus cameras, operable by trained medical assistants (not ophthalmologists), delivering immediate results during the diabetes visit, and automatically routing positive findings to ophthalmology through integrated referral pathways. The system transforms every diabetes clinic into a DR screening center — eliminating the referral-to-specialist bottleneck that is the primary reason half of all diabetic patients are never screened.
A single fundus photograph is a snapshot. A series of fundus photographs over years reveals the trajectory of disease — and that trajectory predicts whether a patient's mild NPDR will remain stable for a decade or progress to proliferative disease within two years. Sentinel Visio analyzes every historical retinal image, quantifies changes in microaneurysm count, hemorrhage area, exudate volume, and vessel caliber over time, and calculates a personalized DR progression risk score. Patients on an accelerating trajectory receive intensified glycemic management and shortened screening intervals — while stable patients can safely extend screening to biennial — personalizing care at scale.
Results from our deployed screening networks.
Sentinel Visio was deployed across 42 primary care clinics serving 68,000 diabetic patients. Before deployment, annual DR screening compliance was 34% — consistent with national averages. Within 18 months, compliance rose to 91%. The system identified 4,200 patients with referable DR who had never been seen by an ophthalmologist, including 340 with vision-threatening disease requiring urgent treatment. Of these 340 patients, 312 received treatment within 30 days of detection. The network's endocrinology director estimated that the program prevented severe vision loss in over 200 patients in its first two years.
A retina practice deployed Sentinel Visio's multi-disease engine across all fundus photographs. Beyond DR grading, the system's glaucoma screening engine flagged 186 patients as glaucoma-suspect from fundus images originally captured for DR screening. IOP measurement and visual field testing confirmed glaucoma in 78 of these patients — 42% — who had no prior diagnosis. The practice's ophthalmologist noted that without the AI glaucoma screening, these patients would have remained undiagnosed until they presented with symptomatic, irreversible visual field loss.
A rural health system with no ophthalmologist within 90 miles deployed Sentinel Visio across 12 community clinics with portable fundus cameras operated by medical assistants. In the first year, 8,400 diabetic patients received point-of-care screening — compared to 1,200 who had made the trip to the nearest ophthalmologist in the prior year. The system identified 620 patients with referable DR and 44 with vision-threatening disease requiring urgent specialist care. Telemedicine referral pathways connected these patients directly with retina specialists for virtual consultation, with in-person visits reserved only for those requiring procedural intervention.
I am an endocrinologist. I manage 2,400 diabetic patients. Before Sentinel Visio, I could beg my patients to see an ophthalmologist, but only a third of them ever went. Now I screen them in my office during their quarterly visit. It takes ninety seconds. The result appears before the patient stands up. I have found vision-threatening retinopathy in patients who told me their eyes were "fine." This is the most impactful technology I have deployed in my practice in twenty years.
The retinal vascular analysis engine found undiagnosed hypertension in a 44-year-old man during a routine DR screen. He had no idea his blood pressure was 168/102. The retina showed arteriovenous nicking and focal arteriolar narrowing — classic grade 2 hypertensive retinopathy. We diagnosed his hypertension from a photograph of his eye. That is the future of screening.
We have twelve clinics spread across a rural county with no ophthalmologist. Before Sentinel Visio, our diabetic patients drove ninety minutes each way for an annual eye exam — and most of them simply didn't go. Now we screen them where they live, where they get their insulin, where they trust their doctor. Screening went from 1,200 to 8,400 patients in one year. That is access. That is equity. That is what AI should do.
Schedule a demonstration of Sentinel Visio — configured for your clinic workflow, your fundus camera, and your patient population.