Modern possibilities for optimizing the calculation of intraocular lens optical power using deep machine learning capabilities

Purpose. To develop optimized machine learning technology to improve the accuracy of intraocular lens (IOL) optical power calculation. Material and methods. Publications in Russian and foreign media fort he last 5 years which report about using artificial intelligence to improve the accuracy of IOL optical power calculation were analyzed. Results. Two different approaches using artificial intelligence in optimization of IOL calculation are known: firstly, modification of existing theoretical formulas and secondly, based only on analysis of patients' database before and after IOL implantation. Each approach has its advantages and disadvantages. The absence of an «ideal» formula forces surgeons to analyze incoming patient data and make corrections to the results obtained based on their own experience when calculating IOLs. In the clinic of the Volgograd branch of the S.N. Fyodorov National Medical Research Center «MNTK «Eye Microsurgery» developed a design project for the LensCalc software application and algorithms of its step-by-step operation. Conclusion. The development of a software application based on artificial intelligence is a rather labor-intensive process that requires processing a large amount of data, using formulas and algorithms with obtaining the final result, as well as subsequent training, optimization, and improving the quality of its work. The LensCalc software application is based on the use of artificial intelligence algorithms to select the formula for calculating the optical power of the IOL that is most suitable in a particular clinical situation. Keywords: IOL calculation, artificial intelligence