Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{190658,
        author = {Shalini R. Bakal and Satish R. Sankaye},
        title = {Comparative Analysis of Clinical Corneal Parameters and Deep CNN Features Derived from Corneal Maps},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {8},
        pages = {1165-1169},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=190658},
        abstract = {The advent of deep convolutional neural networks (CNNs) has revolutionized corneal disease detection, particularly for keratoconus screening. This paper presents a com parative analysis of traditional clinical corneal param eters including K-readings, pachymetry, and elevation maps versus deep CNN derived features extracted from corneal topographic maps. We systematically review recent literature demonstrating that CNN models achieve diagnostic accuracies exceeding 95%, with area under the curve (AUC) values reaching 0.995. While clinical parame ters provide interpretable, standardized measurements, CNN features capture complex spatial patterns that may escape conventional analysis. However, CNN approaches face chal lenges in clinical validation, interpretability, and generalizability. This analysis reveals that hybrid approaches combin ing traditional parameters with CNN derived features, cou pled with explainability methods such as class activation maps, offer the most promising pathway for clinical imple mentation.},
        keywords = {},
        month = {January},
        }