Goal: Leverage patient/doctor time to proactively identify disease risk

The product vision brought to us was to use retina scanning to proactively identify disease risk markers. The underlying idea was to leverage doctor/patient time during regular appointments, without the need for specific procedures.

The concept is relatively simple: a special camera scans the eye and the scan is then processed by a cloud-based service. The results are not insignificant: a reliable indication of disease risk in the patient.

What we achieved:

We created a highly accurate and very handy tool for doctors. It automatically points up factors that may pose a threat to a patient’s health something that is done in a few minutes. Not only is the practicing doctor made aware of potential risk, he also receives recommendations and possible therapy strategies. All this is based on retinal analysis.

How we figured it out:

Crunch the numbers, generate the data, make it available.
Using a 5+ person team, it took us 6 months to deliver.

We still trust in Maths! In order to provide the desired results, we had first to create an engine that can parse around 200 million retinal photos and categorize them. The results of everything that goes on behind the scenes as well as all pertinent information it generates is then made available to doctors via an interface for search filtering. We got familiar with the in-clinic scenario and created an interface that best suits the demands and restrictions of the situation.

The engine producing the service works thanks to an AI service hookup, integration with retinal database and machine learning setup to match retinal diseases with images. We built a mathematical model to reliably find common features based on AI image recognition results.

The tech stack:

Postgress, PHP, JSon, Laravel

Industries:

Healthcare, AI

How’s it all performing?
Using this system led to up to 50% faster diagnoses