The Important Of Data Science In Solving The Myopia Pandemic

By 2050 half of all patients seen by opthalmologists will be myopic, even those patients whose presenting complaints are not related to myopia. If we are to avoid this then we must find methods of research that give useable solutions to prevent and restrict myopia.

Data Science has become a very powerful tool in finding patterns and offering solutions from data that is otherwise incomprehenisble to the human brain.

Unless we harness this resource we are very unlikely to stop and reverse the myopia pandemic

Data harvested from the machines such as biometers, autorefractors, topographers and OCTs is crammed with potential. By mining this data we have a opportunity to interogate it and see what it can tell us.

First find the data. Its in every machine we use and the manufactuerers don't make it obvious because they think we are not interested in it. When extracted, usually as a .CSV file or similar, it can be anonymised, cleaned and preprocessed. It is then ready to be fed into machine learning and intelligent algorithms to show if there are patterns which can help us understand eye growth and the parameters than influence it

In this brief presentation I will show how the data from the Zeiss IOL Master 700 can be extaracted. I will then show how to process it and put it into a machine learing algorithm to learn what it can tell us. One very interesting early result is that the average axial length of a child with myopia is longer than a patient attending for cataract.

As ophthalmologists we are sitting on a mine of data and we don't know it. By encouraging my colleagues to investigate this aspect I hope that we can get closer to a cure for this pandemic.