Predicting the causative agent in microbial keratitis: the role of anterior segment optical coherence tomography-a cross-sectional analysis.

Purpose: To evaluate the anterior segment optical coherence tomography (AS-OCT) features of infectious keratitis (IK) and assess their correlation with clinical and microbiological findings.

Methods: This cross-sectional study included 58 patients with culture-positive corneal abscesses, complete medical records, and high-quality AS-OCT imaging. Independent clinical and AS-OCT assessments were conducted to identify pathogen-specific features and evaluate their diagnostic sensitivity.

Results: Distinct AS-OCT patterns were identified for each pathogen. Gram-positive bacterial keratitis was associated with stromal edema (p 0.006; OR = 20.6) and Descemet folds (p 0.01; OR = 4.83), whereas Gram-negative keratitis exhibited diffuse infiltrates (p < 0.001; OR = 14.286) and stromal thinning (p < 0.001; OR = 98.9). Fungal keratitis was characterized by stromal cystic spaces (p < 0.001; OR = 82.1), stromal thinning (p 0.042; OR = 4.71) and endothelial plaques with indistinct corneal boundaries (p < 0.001; OR = 207). Combining AS-OCT with clinical findings enhanced diagnostic sensitivity to 75% for Gram-positive bacterial keratitis, 66.7% for Gram-negative keratitis, and 100% for fungal keratitis.

Conclusions: AS-OCT provides key pathogen-specific features in IK, improving diagnostic accuracy, particularly in cases with inconclusive microbiological results. While AS-OCT should not replace microbiological testing, its integration into clinical protocols can refine diagnostic strategies. Future applications of artificial intelligence could further optimize early pathogen identification and treatment outcomes.