Abstract
Abstract: :
Purpose: To investigate the relationship between visual field test points in the Humphrey 24-2 test pattern and peripapillary retinal nerve fiber layer (RNFL) thickness measurements estimated with the GDx VCC. Methods: Fifty-one healthy subjects and 91 patients with glaucoma were examined with achromatic automated perimetry (Humphrey Field Analyzer, 24-2 Full Threshold or SITA-Standard test program, Zeiss Humphrey Systems, Dublin, CA, USA) and scanning laser polarimetry with individualized compensation of anterior segment birefringence (GDx VCC, Laser Diagnostic Technologies, Inc., San Diego, CA, USA). The GDx VCC estimates RNFL thickness by measuring the amount of retardation of a polarized laser beam induced by the birefringent RNFL. Individual visual field test points and peripapillary RNFL retardation measurements were grouped into 6 corresponding sectors.1 These sectors were superotemporally (ST), superonasally (SN), nasally (N), inferonasally (IN), inferotemporally (IT), and temporally (T) relative to the optic nerve head. Differential light sensitivity (DLS) was expressed in the typically used logarithmic decibel (dB = 10 * log (1/Lambert)) scale and as 1/Lambert. The relationship between variables was investigated by linear, logarithmic and quadratic regression analysis. A fit with the largest R2 and with coefficients contributing significantly to the equation (P < 0.05), was taken to be significantly best. Results: A linear relationship between 1/Lambert DLS and RNFL thickness was best in all sectors except nasally, in which a logarithmic fit was best. The R2 was lowest in T and N: < 0.25. In the other sectors the R2 varied between 0.31 and 0.56. Between decibel DLS and RNFL thickness, however, a quadratic fit was significantly better than a linear or logarithmic fit for ST, SN, N, and IT. A logarithmic and a linear fit between decibel DLS and RNFL thickness were significantly better for IN and T, respectively. Conclusions: There is a linear relationship between visual function, expressed as 1/Lambert DLS, and peripapillary RNFL thickness estimated with the GDx VCC. A curvilinear relationship exists between decibel DLS and RNFL thickness. 1Garway-Heath DF, Poinoosawmy D, Fitzke FW, Hitchings RA. Ophthalmology. 2000;107:1809-1815.
Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, S • nerve fiber layer • perimetry