Dr. Gairik Kundu

DR. ROHIT SHETTY, DR. SHARON DSOUZA

Abstract

Purpose-To analyse confocal nerve parameters with systemic,orthoptic characteristics in patients with eyepain using Artificial intelligence(AI) model.

Methods-298 eyes of 150 patients with eyepain including 30 patients as controls were analysed.Clinical examination including dry eye tests,systemic,orthoptic evaluation,confocal microscopy(Heidelberg Engineering, Germany)nerve parameters-CNFA,CNFL,CNFD,CNFW,CNFrD with Dendritic Cell(DC) density &microneuromas were evaluated.The subjects were grouped into 4 based on discordance between signs& symptoms.Area under curve (AUC),sensitivity(se),specificity(sp) &accuracy (ac) were evaluated.

Results-AI Model had ac,se,sp of 86%,82.3%&85.6%.AI correctly predicted 81.1%cases in grp1,86.3% in grp 2,69.1% in grp 3&79.7% in grp 4.Important parameters were presence of microneuromas,DCs,orthoptic issues&CNFrD,CNFL.

Conclusion-This demonstrates the importance of the use of AI in stratification of various parameters in patients presenting with eyepain.