AI is the way of the future for IOL power calculation

Storing and analysing high-quality data is becoming faster and cheaper

Priscilla Lynch

Posted: Tuesday, September 17, 2019

Damien Gatinel

Artificial intelligence is improving the precision of intraocular lens (IOL) power calculations and is the future, according to Damien Gatinel MD, France, speaking during the Journal of Cataract and Refractive Surgery Symposium ‘Controversies in Cataract and Refractive Surgery’ session during the 37th Congress of the ESCRS.

Dr Gatinel spoke in favour of using AI technology, specifically deep learning to create adaptive algorithms, for IOL power calculation formulas, adding that we now have access to tremendous amounts of high-quality data and storing and analysing it is becoming faster and cheaper.

Dr Gatinel outlined positive data results on a new AI-based formula that he was involved in developing himself – PEARL-DGS, which was developed using optical and machine learning modelling to predict the effective lens position (ELP) and adjust for extreme biometric values.

The study results for PEARL-DGS compare favourably to classical formulas, and it is good for short and long eyes and complicated cases, including those with previous corneal refractive surgery, he reported.

The more information you get, the more refined the algorithm becomes, Dr Gatinel explained, noting that there was a significant improvement when doing second eye calculations.

Concluding, Dr Gatinel said the IOL calculation community needs a comprehensive “hold-out” data set, independently comparing all existing formulas, across the various eye types and complications, to determine the best options.