Jesper Leth Hougaard1, Anders Heijl, Boel Bengtsson. 1. Department of Clinical Sciences, Ophthalmology, Malmö University Hospital, Lund University, Malmö, Sweden. jesper.hougaard@med.lu.se
Abstract
PURPOSE: We aimed to compare the retinal nerve fibre layer (RNFL) in different Stratus optical coherence tomography (OCT) images in glaucoma eyes with RNFL thickness values within normal limits. METHODS: We studied the Stratus OCT high-resolution standard protocol for assessment of peripapillary RNFL thickness. The four glaucoma eyes presented (with reproducible visual field defects, mean deviations from -3.49 dB to -0.49 dB and structural loss on RNFL and/or disc photography) had results within the 95% normal limits for all OCT parameters (full circle, quadrant or clock hour mean RNFL thickness). We qualitatively compared the RNFL in pseudo-colour and grey-scale images defined by the Stratus OCT software with what appeared to be the RNFL when not defined automatically (and in raw data OCT images). RESULTS: A localized RNFL defect was identified in conventional 495-nm RNFL photographs in all four patients. Obviously, in these selected eyes, the atrophy was not picked up by the OCT RNFL border algorithm, and at the corresponding location a localized RNFL atrophy (narrowing of the high signal intensity RNFL) was either not or just barely detectable in pseudo-colour OCT. However, defects seemed to be present in OCT grey-scale and raw data images in all four eyes. CONCLUSIONS: Our observations suggest that OCT images obtained in glaucoma eyes that are misclassified by the standard statistical analyses of Stratus OCT may contain information indicating RNFL defects. Thus, the current diagnostic ability of OCT in glaucoma can probably be extended.
PURPOSE: We aimed to compare the retinal nerve fibre layer (RNFL) in different Stratus optical coherence tomography (OCT) images in glaucoma eyes with RNFL thickness values within normal limits. METHODS: We studied the Stratus OCT high-resolution standard protocol for assessment of peripapillary RNFL thickness. The four glaucoma eyes presented (with reproducible visual field defects, mean deviations from -3.49 dB to -0.49 dB and structural loss on RNFL and/or disc photography) had results within the 95% normal limits for all OCT parameters (full circle, quadrant or clock hour mean RNFL thickness). We qualitatively compared the RNFL in pseudo-colour and grey-scale images defined by the Stratus OCT software with what appeared to be the RNFL when not defined automatically (and in raw data OCT images). RESULTS: A localized RNFL defect was identified in conventional 495-nm RNFL photographs in all four patients. Obviously, in these selected eyes, the atrophy was not picked up by the OCT RNFL border algorithm, and at the corresponding location a localized RNFL atrophy (narrowing of the high signal intensity RNFL) was either not or just barely detectable in pseudo-colour OCT. However, defects seemed to be present in OCT grey-scale and raw data images in all four eyes. CONCLUSIONS: Our observations suggest that OCT images obtained in glaucoma eyes that are misclassified by the standard statistical analyses of Stratus OCT may contain information indicating RNFL defects. Thus, the current diagnostic ability of OCT in glaucoma can probably be extended.