How it Works

Smartphone/App

OcuDoc has developed a proprietary algorithm that analyzes light reflected from the retina. This “scatter pattern” is detected by the sensors on a smartphone and sent to our proprietary and secure cloud. Using machine learning and artificial intelligence, we are able to determine a change in axial length (length of the eyeball). If axial length gets longer, myopia (nearsightedness) gets worse.

This is revolutionary because now there is a truly mobile way of measuring changes in axial length and refraction. OcuDoc uses your smartphone, so there is no expensive machine to buy. It is convenient and accurate. Quick evaluations can be made to determine if you need to see an eye care professional for changes in axial length or your prescription has changed.

This same technology also gives the strong possibility of detecting retinal detachment and neurologic conditions with eye movement components.

Smartphone Sensors

This technology is enabled by the sensors on smartphones. Sensors detect light, which is how you take photos with your smartphone. Instead of the light that hits the sensors turning into a photo, the "raw" sensor data is sent directly to our cloud. From there, our proprietary algorithm does the rest. The sensor data is NOT analyzed by the smartphone.

Data

In order to develop these capabilities, we have scanned thousands of human eyes (living people) and compared the data to the data from reference devices. A reference device is an existing machine that optometrists and ophthalmologists use to measure axial length and for other ocular examinations. These machines usually cost more than $25,000. We have achieved a level of accuracy that matches and sometimes exceeds these devices. The two devices we compare with are the Topcon Myah and the Zeiss IOL Master.

AI

Artificial Intelligence is “the ability of a machine to perform tasks that are usually associated with human intelligence, such as learning, reasoning, and making decisions.” With our technology, our AI gets the sensor data from a new person, compares it to all the eye scans from our database of eye scans, and then tells us the current axial length and the change in axial length, or eyeglass prescription. It takes thousands of eye scans to build a database. In addition, the AI must be “trained”. In the beginning, the AI must be “taught” all the different human eye variables, such as, “this is a blue eye, this is a green eye, these are thick eye lashes, etc.” After this labor intensive process, the AI can start “learning” on its own.