ESCRS - POANT014 - Corneal Densitometry Maps For Diagnosis Of Common Corneal Diseases: A Multigrader Analysis

Corneal Densitometry Maps For Diagnosis Of Common Corneal Diseases: A Multigrader Analysis

Published 2026 - 44th Congress of the ESCRS

Reference: POANT014 | Type: Poster | DOI: 10.82333/2bw4-gm53

Authors: Mr Elias Golschan* 1 , Dr Titus Schug 1 , Dr Tyll Jandewerth 1 , Dr Loay Daas 2 , Dr Ingo Schmack 1 , A/Prof. Thomas Kohnen 1 , Dr Klemens Paul Kaiser 1

1Ophthalmology,Goethe University,Frankfurt,Germany, 2Ophthalmology,University of Saarland,Homburg Saar,Germany

Purpose

To evaluate the diagnostic performance of corneal optical densitometry (COD) maps for discrimination of common corneal diseases in a masked multigrader setting.

Setting

Retrospective image-based diagnostic accuracy study with prospective masked reader evaluation.
Tertiary referral center, Department of Ophthalmology, Goethe University Frankfurt, Germany.

Methods

Five hundred corneal optical densitometry maps obtained with Scheimpflug imaging were identified retrospectively. After exclusion of 50 images used for training, 450 maps (90 healthy controls, 90 dry eye disease (DED), 90 keratoconus (KCN), 90 Fuchs endothelial corneal dystrophy (FECD), 90 epithelial basement membrane dystrophy (EBMD) were independently graded by four masked observers with different experience levels. Diagnostic accuracy was assessed using a one-versus-rest approach and expressed as sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Intergrader agreement was evaluated using Fleiss’ kappa.

Results

Overall diagnostic accuracy ranged from 68.0 % (resident) to 77.1 % (senior consultant). Intergrader agreement was substantial (κ = 0.63). Sensitivity varied across entities, ranging from 47.8 % for DED to 95.6 % for FECD, whereas specificity remained consistently high (86.7 %- 98.9 %). Diagnostic performance was highest for FECD and EBMD. Most misclassifications occurred between healthy controls and DED. Accuracy differed significantly between graders (p < 0.001), demonstrating an association between experience level and correct classification.

 

Conclusions

COD maps allow moderate to high diagnostic discrimination of common corneal diseases, particularly in structurally defined conditions such as FECD and EBMD. While densitometry does not replace clinical examination, it may provide incremental value as an adjunctive structural imaging tool in corneal diagnostics.