PLoS ONE · 2016 · 22 citations · 18 references
MedicineOphthalmologyVision ScienceOcular ImagingOcular PathologyVisual OpticsCataract AssessmentCataract EvaluationLens PathologyOphthalmic ImagingCataract ClassificationCortical Cataract AssessmentCortical Cataract GradingLens Opacity GradingNuclear Cataract AssessmentNuclear Cataract GradingOpacityOpacity Classification
PURPOSE: To assess nuclear and cortical opacities through the objective analysis of Scheimpflug images, and to check the correlation with the Lens Opacity Classification System III (LOCS III). METHODS: Nuclear and cortical opacities were graded according to the LOCS III rules after pupil dilation. The maximum and average pixel intensity values along an elliptical mask within the lens nucleus were taken to analyse nuclear cataracts. A new metric based on the percentage of opaque pixels within a region of interest was used to analyse cortical cataracts. The percentage of opaque pixels was also calculated for half, third and quarter areas from the region of interest's periphery. RESULTS: The maximum and average intensity values along the nucleus were directly proportional to the LOCS III grade: The larger the LOCS III value, the larger maximum and average intensity ones. These metrics showed a positive and significant correlation with the LOCS grade: The larger the LOCS grade, the higher was percentage of opaque pixels along the cortex within the same mask's size. This metric showed a significant correlation to the LOCS grade. CONCLUSION: The metrics used to assess nuclear opacities showed good correlation with the LOCS III. The percentage of opaque pixels showed to be a useful metric to measure objectively the severity of the cortical opacity. These metrics could be implemented in an algorithm to detect and grade lens opacities automatically and objectively.
18
The Lens Opacities Classification System III
L.T. Chylack · Archives of Ophthalmology · 1993 · 2.9K citations
Xiaomei Pei, Yongzhen Bao, Y Chen et al. · British Journal of Ophthalmology · 2008 · 126 citations
A Computer-Aided Diagnosis System of Nuclear Cataract
Huiqi Li, Joo‐Hwee Lim, Jiang Liu et al. · IEEE Transactions on Biomedical Engineering · 2010 · 104 citations