PURPOSE: To determine whether neural network techniques can improve differentiation between glaucomatous and nonglaucomatous eyes, using the optic disc topography parameters of the Heidelberg Retina Tomograph (HRT; Heidelberg Engineering, Heidelberg, Germany). METHODS: With the HRT, one eye was imaged from each of 108 patients with glaucoma (defined as having repeatable visual field defects with standard automated perimetry) and 189 subjects without glaucoma (no visual field defects with healthy-appearing optic disc and retinal nerve fiber layer on clinical examination) and the optic nerve topography was defined by 17 global and 66 regional HRT parameters. With all the HRT parameters used as input, receiver operating characteristic (ROC) curves were generated for the classification of eyes, by three neural network techniques: linear and Gaussian support vector machines (SVM linear and SVM Gaussian, respectively) and a multilayer perceptron (MLP), as well as four previously proposed linear discriminant functions (LDFs) and one LDF developed on the current data with all HRT parameters used as input. RESULTS: The areas under the ROC curves for SVM linear and SVM Gaussian were 0.938 and 0.945, respectively; for MLP, 0.941; for the current LDF, 0.906; and for the best previously proposed LDF, 0.890. With the use of forward selection and backward elimination optimization techniques, the areas under the ROC curves for SVM Gaussian and the current LDF were increased to approximately 0.96. CONCLUSIONS: Trained neural networks, with global and regional HRT parameters used as input, improve on previously proposed HRT parameter-based LDFs for discriminating between glaucomatous and nonglaucomatous eyes. The performance of both neural networks and LDFs can be improved with optimization of the features in the input. Neural network analyses show promise for increasing diagnostic accuracy of tests for glaucoma.
PURPOSE: To determine whether neural network techniques can improve differentiation between glaucomatous and nonglaucomatous eyes, using the optic disc topography parameters of the Heidelberg Retina Tomograph (HRT; Heidelberg Engineering, Heidelberg, Germany). METHODS: With the HRT, one eye was imaged from each of 108 patients with glaucoma (defined as having repeatable visual field defects with standard automated perimetry) and 189 subjects without glaucoma (no visual field defects with healthy-appearing optic disc and retinal nerve fiber layer on clinical examination) and the optic nerve topography was defined by 17 global and 66 regional HRT parameters. With all the HRT parameters used as input, receiver operating characteristic (ROC) curves were generated for the classification of eyes, by three neural network techniques: linear and Gaussian support vector machines (SVM linear and SVM Gaussian, respectively) and a multilayer perceptron (MLP), as well as four previously proposed linear discriminant functions (LDFs) and one LDF developed on the current data with all HRT parameters used as input. RESULTS: The areas under the ROC curves for SVM linear and SVM Gaussian were 0.938 and 0.945, respectively; for MLP, 0.941; for the current LDF, 0.906; and for the best previously proposed LDF, 0.890. With the use of forward selection and backward elimination optimization techniques, the areas under the ROC curves for SVM Gaussian and the current LDF were increased to approximately 0.96. CONCLUSIONS: Trained neural networks, with global and regional HRT parameters used as input, improve on previously proposed HRT parameter-based LDFs for discriminating between glaucomatous and nonglaucomatous eyes. The performance of both neural networks and LDFs can be improved with optimization of the features in the input. Neural network analyses show promise for increasing diagnostic accuracy of tests for glaucoma.
Authors: Michael H Goldbaum; Pamela A Sample; Zuohua Zhang; Kwokleung Chan; Jiucang Hao; Te-Won Lee; Catherine Boden; Christopher Bowd; Rupert Bourne; Linda Zangwill; Terrence Sejnowski; David Spinak; Robert N Weinreb Journal: Invest Ophthalmol Vis Sci Date: 2005-10 Impact factor: 4.799
Authors: Christopher Bowd; Intae Lee; Michael H Goldbaum; Madhusudhanan Balasubramanian; Felipe A Medeiros; Linda M Zangwill; Christopher A Girkin; Jeffrey M Liebmann; Robert N Weinreb Journal: Invest Ophthalmol Vis Sci Date: 2012-04-30 Impact factor: 4.799
Authors: Zvia Burgansky-Eliash; Gadi Wollstein; Tianjiao Chu; Joseph D Ramsey; Clark Glymour; Robert J Noecker; Hiroshi Ishikawa; Joel S Schuman Journal: Invest Ophthalmol Vis Sci Date: 2005-11 Impact factor: 4.799
Authors: Michael H Goldbaum; Irina Falkenstein; Igor Kozak; Jiucang Hao; Dirk-Uwe Bartsch; Terrance Sejnowski; William R Freeman Journal: Trans Am Ophthalmol Soc Date: 2008
Authors: Igor Kozak; Pamela A Sample; Jiucang Hao; William R Freeman; Robert N Weinreb; Te-Won Lee; Michael H Goldbaum Journal: Trans Am Ophthalmol Soc Date: 2007
Authors: K A Townsend; G Wollstein; D Danks; K R Sung; H Ishikawa; L Kagemann; M L Gabriele; J S Schuman Journal: Br J Ophthalmol Date: 2008-06 Impact factor: 4.638