Abstract
Purpose: :
We demonstrated that, in patients with severe myopia, the distribution of retinal nerve fiber layer (RNFL) thickness shifted to more temporal. Anew Fourier–based analysis method was evaluated for diagnosing glaucoma in patients with severe myopia. To improve diagnostic capability, subjects were divided based on myopic refraction and we also evaluated the new analysis method.
Methods: :
RNFL thickness estimates were obtained from 40 healthy individuals and 40 glaucoma patients with myopia of –5 diopters and greater. All subjects were imaged using a scanning laser polarimeter (SLP) with variable corneal compensation (GDx VCC, Laser Diagnostic Technology) and optical coherence tomography (OCT 3000, Zeiss).The pattern of thickness measurements from SLP and OCT in the 32 sectors around the optic disc was analyzed to obtain the Fourier coefficients. The fast Fourier transformation (FFT) was employed to determine the coefficients. The values were entered into linear discriminant analysis. Receiver operating characteristic (ROC) curve were used to compare the performance of the Fourier–based metrics against other commonly used RNFL analytical procedures.
Results: :
The area under the ROC curve (AUC) using the linear discriminant function (LDF)based on Fourier analysis obtained by OCT and SLP was 0.891 and 0.878, respectively, and greater compared with any other analytical procedures, including GDx number (0.812). Healthy subjects and glaucoma patients were divided based on myopic refraction into two groups; –5 to –7.9 diopters and –8 diopters and higher. In eyes with –5 to –7.9 diopters, the AUC using Fourier– based LDF obtained by OCT and SLP was 0.946 and 0.934, respectively. In eyes with –8.0D and greater, the AUC using Fourier– based LDF obtained by OCT and SLP was 0.908 and 0.915. In eyes with –5 to –7.9 diopters and –8 diopters and higher, diagnostic capability of the Fourier– based LDF obtained by the two systems was better than that of any other analytical procedures.
Conclusions: :
In patients with severe myopia, the discriminant function based on the output from a Fourier analysis of RNFL data resulted in better diagnostic capability compared with other common RNFL analytical procedures. Diagnostic capability improved when subjects were divided based on myopic refraction into two groups. No significant difference was found between OCT and SLP.
Keywords: imaging/image analysis: clinical • myopia • nerve fiber layer