AI that sees ROP
before it's too late.

ROP is a potentially blinding eye disorder that primarily affects premature infants weighing about 2¾ pounds (1250 grams) or less who are born before 31 weeks of gestation. It is caused by abnormal blood vessel growth in the light-sensitive retina of the eye.

Blood vessels in the retina normally complete their development a few weeks before birth. In premature infants, this process is disrupted. The eye then tries to repair itself by growing new vessels, but these are often abnormal and can lead to scarring and retinal detachment. Key risk factors include low birth weight, early gestational age, and supplemental oxygen therapy.

NeoVuo AI acts as an advanced triage tool. While a standard diagnosis relies solely on the ophthalmologist's manual review at the bedside, our AI analyzes retinal images to rapidly flag high-risk zones (Zone I/II) and 'Plus' disease, ensuring that urgent cases are prioritized for the doctor's attention.

Not all babies with ROP need treatment. In many mild cases (Stage 1 and 2), the condition resolves on its own as the eye matures. Treatment is usually required only for severe cases (Stage 3 with Plus disease or higher) to prevent retinal detachment.

ROP is classified into 5 stages: Stage 1 (mildly abnormal vessel growth), Stage 2 (moderately abnormal), Stage 3 (severely abnormal vessel growth), Stage 4 (partial retinal detachment), and Stage 5 (total retinal detachment). 'Plus disease' indicates a severe, rapidly progressing form.

Yes. NeoVuo AI is a software tool that processes images taken during standard eye exams. It does not require any additional physical contact or procedures with the infant. The original eye exam is conducted by trained medical staff using standard protocols.

Infants with severe ROP are at higher risk for vision problems later, such as nearsightedness (myopia), strabismus (crossed eyes), or amblyopia (lazy eye). Regular follow-up exams with an ophthalmologist are crucial even after the ROP has resolved.

No. NeoVuo AI is an assistive triage tool. It highlights potential risk areas and suggests staging to speed up the screening process, but the final diagnosis and treatment decision always remains with the qualified clinician.

Our models are trained on clinically validated datasets (like KIDROP and proprietary local datasets). While accuracy is high in testing environments, we are currently in the research demo phase to validate real-world performance.

Security is our priority. We use an anonymization pipeline that strips PII (Personal Identifiable Information) from images before processing. For the demo, we operate on a 'Zero-Retention' policy, meaning images are deleted after analysis.

Access is currently restricted to research partners and clinical observers. You can request access by filling out the 'Request Research Demo Access' form below to join our waitlist.