Use of artificial intelligence to evaluate the detection of alterations in the retina as a screening test in Mexican patients

Abstract

In Mexico, chronic degenerative diseases are the main cause of morbidity, which have frequent complications in the retina, being the main cause of blindness in our population. Unfortunately, the detection of the pathology is usually late, which causes greater disability. To propose the detection of different pathologies, different artificial intelligence algorithms have been used for the images taken from the fundus. Objective. To evaluate different machine learning algorithms for screening of retinal alterations in the Mexican population. Methodology. We evaluate two types of models to estimate the screening capacity of artificial intelligence tools, one based on transfer learning, and ensemble methods against one based solely on convolutional networks. Results. We obtained adequate values to differentiate between healthy and sick, but not to diagnose different pathologies. Conclusion. It is necessary to expand the sample of images and improve the screening models