Detection of glaucomatous retinal nerve fiber layer damage by scanning laser polarimetry with variable corneal compensation

摘要

One of the earliest signs of glaucoma presence is defects in the retinal nerve fiber layer (RNFL). Scanning laser polarimetry (SLP) provides objective assessment of RNFL, a birefringent tissue, by measuring the total retardation in the reflected light. SLP provides a potential tool for early detection of glaucoma and its progression. The birefringence of the anterior segment of the eye, mainly the cornea, is a confounding variable to SLP's clinical application, if compensation cannot be achieved properly. This paper presents a new SLP system, GDx VCC (Laser Diagnostic Technologies, Inc., San Diego, CA), with a variable corneal compensator (VCC) to achieve individualized corneal compensation. Clinical application of this device in glaucoma detection is also demonstrated. Methods: The GDx VCC system is a confocal scanning laser ophthalmoscope integrated with an ellipsometer. Image field is 40° (H) by 20° (V), covering both the peripapillary region and the macular region of the eye. The variable corneal compensator consists of two identical linear retarders; corneal birefringence is measured from the SLP macular image; both the magnitude and axis of the VCC are adjusted to the required compensation and RNFL measurement is performed. Clinical data were collected from 5 sites in the US under one study protocol; RNFL images of 390 normal eyes and 253 glaucomatous eyes were acquired with GDx. VCC systems. A normative database and a neural network classifier were developed so that RNFL status can be assessed objectively even at first exam. Results: GDx VCC generates 4 images for each measurement: fundus reflectance image, retardation image, axis image, and depolarized light image. Corneal birefringence magnitude and axis both varied over wide ranges for the study population, confirming the necessity of VCC. The VCC method and device as implemented in the GDx VCC systems worked well, apparent from the compensated macular retardation images. Glaucomatous RNFL damage, both focal defects and diffuse defects, was readily identified in the retardation images. RNFL measurement parameters with normative database provided objective RNFL evaluation consistent with clinical diagnosis. RNFL damage prior to visual field loss was observed in eyes diagnosed with early stage glaucoma. Significant correlation was observed between RNFL measurements and visual field mean deviation for the patient group. Conclusions: Individualized corneal compensation is achieved with the GDx VCC system. Glaucomatous RNFL damage can be detected effectively with GDx VCC technology. Normative database and neural network classifier allow diagnosis at a single exam. The significant association between RNFL measurement and visual field mean deviation in glaucomatous eyes suggests that GDx VCC may be a tool for monitoring progression of the disease.