Comparing neural networks and linear discriminant functions for glaucoma detection using confocal scanning laser ophthalmoscopy of the optic disc.

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.

[1]  R N Weinreb,et al.  Effect of repetitive imaging on topographic measurements of the optic nerve head. , 1993, Archives of ophthalmology.

[2]  L Brigatti,et al.  Neural networks to identify glaucoma with structural and functional measurements. , 1996, American journal of ophthalmology.

[3]  Robert N Weinreb,et al.  Comparing machine learning classifiers for diagnosing glaucoma from standard automated perimetry. , 2002, Investigative ophthalmology & visual science.

[4]  S M Drance,et al.  The ability of scanning laser ophthalmoscopy to identify various glaucomatous optic disk appearances. , 1998, American journal of ophthalmology.

[5]  Volker Roth,et al.  Nonlinear Discriminant Analysis Using Kernel Functions , 1999, NIPS.

[6]  L. Zangwill,et al.  Relationship of optic disc topography to optic nerve fiber number in glaucoma. , 1998, Archives of ophthalmology.

[7]  D. Broomhead,et al.  Radial Basis Functions, Multi-Variable Functional Interpolation and Adaptive Networks , 1988 .

[8]  Robert N Weinreb,et al.  Imaging of the optic disc and retinal nerve fiber layer: the effects of age, optic disc area, refractive error, and gender. , 2002, Journal of the Optical Society of America. A, Optics, image science, and vision.

[9]  J. Jonas,et al.  Discriminant analysis models for early detection of glaucomatous optic disc changes , 2000, The British journal of ophthalmology.

[10]  S M Drance,et al.  ROC analysis of Heidelberg Retina Tomograph optic disc shape measures in glaucoma. , 1997, Canadian journal of ophthalmology. Journal canadien d'ophtalmologie.

[11]  L. Zangwill,et al.  Discriminating between normal and glaucomatous eyes using the Heidelberg Retina Tomograph, GDx Nerve Fiber Analyzer, and Optical Coherence Tomograph. , 2001, Archives of ophthalmology.

[12]  L Brigatti,et al.  Automatic detection of glaucomatous visual field progression with neural networks. , 1997, Archives of ophthalmology.

[13]  E. DeLong,et al.  Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. , 1988, Biometrics.

[14]  Heekuck Oh,et al.  Neural Networks for Pattern Recognition , 1993, Adv. Comput..

[15]  J Katz,et al.  Neural networks for visual field analysis: how do they compare with other algorithms? , 1999, Journal of glaucoma.

[16]  M H Goldbaum,et al.  Interpretation of automated perimetry for glaucoma by neural network. , 1994, Investigative ophthalmology & visual science.

[17]  Nello Cristianini,et al.  Large Margin DAGs for Multiclass Classification , 1999, NIPS.

[18]  M Schulzer,et al.  Reproducibility of topographic parameters obtained with the heidelberg retina tomograph. , 1993, Journal of glaucoma.

[19]  K Rohrschneider,et al.  Reproducibility of the optic nerve head topography with a new laser tomographic scanning device. , 1994, Ophthalmology.

[20]  Vladimir Cherkassky,et al.  The Nature Of Statistical Learning Theory , 1997, IEEE Trans. Neural Networks.

[21]  Yoshua Bengio,et al.  Pattern Recognition and Neural Networks , 1995 .

[22]  N. Swindale,et al.  Ability of the Heidelberg Retina Tomograph to Detect Early Glaucomatous Visual Field Loss , 1995, Journal of glaucoma.

[23]  J. Caprioli,et al.  Detection of structural damage from glaucoma with confocal laser image analysis. , 1996, Investigative ophthalmology & visual science.

[24]  J. Funk,et al.  Optic nerve head analyser and Heidelberg retina tomograph: accuracy and reproducibility of topographic measurements in a model eye and in volunteers. , 1994, The British journal of ophthalmology.

[25]  G. Wollstein,et al.  Identification of early glaucoma cases with the scanning laser ophthalmoscope. , 1998, Ophthalmology.

[26]  Jost B Jonas,et al.  Preperimetric glaucoma diagnosis by confocal scanning laser tomography of the optic disc , 1999, The British journal of ophthalmology.

[27]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[28]  Vladimir Vapnik,et al.  Statistical learning theory , 1998 .

[29]  David S. Broomhead,et al.  Multivariable Functional Interpolation and Adaptive Networks , 1988, Complex Syst..

[30]  David Keating,et al.  Visual field interpretation with a personal computer based neural network , 1994, Eye.

[31]  D B Henson,et al.  Visual field analysis using artificial neural networks , 1994, Ophthalmic & physiological optics : the journal of the British College of Ophthalmic Opticians.

[32]  Geoffrey E. Hinton,et al.  Learning representations of back-propagation errors , 1986 .

[33]  J Caprioli,et al.  Slope of the peripapillary nerve fiber layer surface in glaucoma. , 1998, Investigative ophthalmology & visual science.

[34]  B C Chauhan,et al.  Test-retest variability of topographic measurements with confocal scanning laser tomography in patients with glaucoma and control subjects. , 1994, American journal of ophthalmology.

[35]  P A Sample,et al.  Detection of early glaucomatous structural damage with confocal scanning laser tomography. , 1998, Journal of glaucoma.

[36]  John C. Platt,et al.  Fast training of support vector machines using sequential minimal optimization, advances in kernel methods , 1999 .

[37]  Christopher Bowd,et al.  Evaluating the Optic Disc and Retinal Nerve Fiber Layer in Glaucoma II: Optical Image Analysis , 2000, Seminars in ophthalmology.