Haris Sideroudi1, Georgios Labiris2, Kimon Georgatzoglou2, Fienke Ditzel2, Charalambos Siganos2, Vassilios Kozobolis2. 1. From the Eye Institute of Thrace (Sideroudi, Georgios, Georgatzoglou, Vassilios), Democritus University, and the Department of Ophthalmology (Georgios, Vassilios), University Hospital of Alexandroupolis, and the Institute of Vision and Optics (Signos), University of Crete, Heraklion, Greece; the Medical Center (Ditzel), University of Groningen, Groningen, the Netherlands. Electronic address: harissid@alex.duth.gr. 2. From the Eye Institute of Thrace (Sideroudi, Georgios, Georgatzoglou, Vassilios), Democritus University, and the Department of Ophthalmology (Georgios, Vassilios), University Hospital of Alexandroupolis, and the Institute of Vision and Optics (Signos), University of Crete, Heraklion, Greece; the Medical Center (Ditzel), University of Groningen, Groningen, the Netherlands.
Abstract
PURPOSE: To evaluate the contribution of Fourier analysis of videokeratographic data in the diagnosis of subclinical keratoconus and keratoconus. SETTING: Eye Institute of Thrace, Democritus University, Alexandroupolis, Greece. DESIGN: Observational case series. METHODS: The following Pentacam-derived parameters, resulting from Fourier decomposition of keratometric data, were evaluated for their diagnostic capacity using receiver operating curves: spherical component and eccentricity, maximum decentration, regular astigmatism in the center and in the periphery, mean astigmatism, irregularities, regular astigmatism in the center plus the irregularities, and total astigmatism. Logistic regression was performed to identify a combined diagnostic model. RESULTS: The study comprised 80 keratoconus eyes, 55 eyes diagnosed with subclinical keratoconus, and 50 normal eyes. Significant differences were detected in spherical eccentricity, maximum decentration, irregularities, regular astigmatism in the center and in the periphery, regular astigmatism in the center plus the irregularities, mean astigmatism, and total astigmatism parameters between the groups. Almost all parameters had high diagnostic ability in both study groups (area under the curve >90%). Among individual parameters, those with the highest predictive accuracy were the regular astigmatism in the center plus the irregularities (subclinical keratoconus 97.6%, keratoconus 98.8%) and the maximum decentration (subclinical keratoconus 91.4%, keratoconus 98.5%). Sufficient predictive accuracy (subclinical keratoconus 99.4, keratoconus 100%) was identified in a diagnostic model that combined the regular astigmatism in the center plus the irregularities and the maximum decentration. CONCLUSION: Fourier decomposition of keratometric data provided parameters with high accuracy in differentiating corneas with subclinical keratoconus from normal corneas and should be included to allow prompt diagnosis of keratoconus. FINANCIAL DISCLOSURE: None of the authors has a financial or proprietary interest in any material or method mentioned.
PURPOSE: To evaluate the contribution of Fourier analysis of videokeratographic data in the diagnosis of subclinical keratoconus and keratoconus. SETTING: Eye Institute of Thrace, Democritus University, Alexandroupolis, Greece. DESIGN: Observational case series. METHODS: The following Pentacam-derived parameters, resulting from Fourier decomposition of keratometric data, were evaluated for their diagnostic capacity using receiver operating curves: spherical component and eccentricity, maximum decentration, regular astigmatism in the center and in the periphery, mean astigmatism, irregularities, regular astigmatism in the center plus the irregularities, and total astigmatism. Logistic regression was performed to identify a combined diagnostic model. RESULTS: The study comprised 80 keratoconus eyes, 55 eyes diagnosed with subclinical keratoconus, and 50 normal eyes. Significant differences were detected in spherical eccentricity, maximum decentration, irregularities, regular astigmatism in the center and in the periphery, regular astigmatism in the center plus the irregularities, mean astigmatism, and total astigmatism parameters between the groups. Almost all parameters had high diagnostic ability in both study groups (area under the curve >90%). Among individual parameters, those with the highest predictive accuracy were the regular astigmatism in the center plus the irregularities (subclinical keratoconus 97.6%, keratoconus 98.8%) and the maximum decentration (subclinical keratoconus 91.4%, keratoconus 98.5%). Sufficient predictive accuracy (subclinical keratoconus 99.4, keratoconus 100%) was identified in a diagnostic model that combined the regular astigmatism in the center plus the irregularities and the maximum decentration. CONCLUSION: Fourier decomposition of keratometric data provided parameters with high accuracy in differentiating corneas with subclinical keratoconus from normal corneas and should be included to allow prompt diagnosis of keratoconus. FINANCIAL DISCLOSURE: None of the authors has a financial or proprietary interest in any material or method mentioned.