Relevance vector machine and support vector machine classifier analysis of scanning laser polarimetry retinal nerve fiber layer measurements.

PURPOSE To classify healthy and glaucomatous eyes using relevance vector machine (RVM) and support vector machine (SVM) learning classifiers trained on retinal nerve fiber layer (RNFL) thickness measurements obtained by scanning laser polarimetry (SLP). METHODS Seventy-two eyes of 72 healthy control subjects (average age = 64.3 +/- 8.8 years, visual field mean deviation = -0.71 +/- 1.2 dB) and 92 eyes of 92 patients with glaucoma (average age = 66.9 +/- 8.9 years, visual field mean deviation = -5.32 +/- 4.0 dB) were imaged with SLP with variable corneal compensation (GDx VCC; Laser Diagnostic Technologies, San Diego, CA). RVM and SVM learning classifiers were trained and tested on SLP-determined RNFL thickness measurements from 14 standard parameters and 64 sectors (approximately 5.6 degrees each) obtained in the circumpapillary area under the instrument-defined measurement ellipse (total 78 parameters). Ten-fold cross-validation was used to train and test RVM and SVM classifiers on unique subsets of the full 164-eye data set and areas under the receiver operating characteristic (AUROC) curve for the classification of eyes in the test set were generated. AUROC curve results from RVM and SVM were compared to those for 14 SLP software-generated global and regional RNFL thickness parameters. Also reported was the AUROC curve for the GDx VCC software-generated nerve fiber indicator (NFI). RESULTS The AUROC curves for RVM and SVM were 0.90 and 0.91, respectively, and increased to 0.93 and 0.94 when the training sets were optimized with sequential forward and backward selection (resulting in reduced dimensional data sets). AUROC curves for optimized RVM and SVM were significantly larger than those for all individual SLP parameters. The AUROC curve for the NFI was 0.87. CONCLUSIONS Results from RVM and SVM trained on SLP RNFL thickness measurements are similar and provide accurate classification of glaucomatous and healthy eyes. RVM may be preferable to SVM, because it provides a Bayesian-derived probability of glaucoma as an output. These results suggest that these machine learning classifiers show good potential for glaucoma diagnosis.

[1]  Andrew P. Sage,et al.  Uncertainty in Artificial Intelligence , 1987, IEEE Transactions on Systems, Man, and Cybernetics.

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

[3]  R. Weinreb,et al.  Histopathologic validation of Fourier-ellipsometry measurements of retinal nerve fiber layer thickness. , 1990, Archives of ophthalmology.

[4]  Douglas R. Anderson,et al.  Clinical Decisions In Glaucoma , 1993 .

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

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

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

[8]  R. Weinreb Evaluating the retinal nerve fiber layer in glaucoma with scanning laser polarimetry. , 1999, Archives of ophthalmology.

[9]  Christopher M. Bishop,et al.  Variational Relevance Vector Machines , 2000, UAI.

[10]  R. Knighton,et al.  Effect of corneal polarization axis on assessment of retinal nerve fiber layer thickness by scanning laser polarimetry. , 2000, American journal of ophthalmology.

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

[12]  L. Zangwill,et al.  Detecting early glaucoma by assessment of retinal nerve fiber layer thickness and visual function. , 2001, Investigative ophthalmology & visual science.

[13]  H. Lemij,et al.  Prevalence of split nerve fiber layer bundles in healthy eyes imaged with scanning laser polarimetry. , 2001, Ophthalmology.

[14]  George Eastman House,et al.  Sparse Bayesian Learning and the Relevance Vector Machine , 2001 .

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

[16]  Robert N Weinreb,et al.  Using machine learning classifiers to identify glaucomatous change earlier in standard visual fields. , 2002, Investigative ophthalmology & visual science.

[17]  Joseph Caprioli,et al.  Correction for the erroneous compensation of anterior segment birefringence with the scanning laser polarimeter for glaucoma diagnosis. , 2002, Investigative ophthalmology & visual science.

[18]  Robert N Weinreb,et al.  Comparing neural networks and linear discriminant functions for glaucoma detection using confocal scanning laser ophthalmoscopy of the optic disc. , 2002, Investigative ophthalmology & visual science.

[19]  R. Weinreb,et al.  Individualized compensation of anterior segment birefringence during scanning laser polarimetry. , 2002, Investigative ophthalmology & visual science.

[20]  L. Zangwill,et al.  Measurement of the magnitude and axis of corneal polarization with scanning laser polarimetry. , 2002, Archives of ophthalmology.

[21]  Bagging Tree Classifiers for Laser Scanning Images: Data and Simulation Based Strategy , 2002, Artif. Intell. Medicine.

[22]  Christopher M. Bishop,et al.  Bayesian Regression and Classification , 2003 .

[23]  Robert N Weinreb,et al.  Association between scanning laser polarimetry measurements using variable corneal polarization compensation and visual field sensitivity in glaucomatous eyes. , 2003, Archives of ophthalmology.

[24]  Torsten Hothorn,et al.  New glaucoma classification method based on standard Heidelberg Retina Tomograph parameters by bagging classification trees. , 2003, Journal of glaucoma.

[25]  F A Medeiros,et al.  Comparison of algorithms for detection of localised nerve fibre layer defects using scanning laser polarimetry , 2003, The British journal of ophthalmology.

[26]  L. Zangwill,et al.  Glaucoma detection using scanning laser polarimetry with variable corneal polarization compensation. , 2003, Archives of ophthalmology.

[27]  W. Feuer,et al.  Scanning laser polarimetry with variable corneal compensation and optical coherence tomography in normal and glaucomatous eyes. , 2003, American journal of ophthalmology.

[28]  F. Medeiros,et al.  Optic Nerve Imaging: Recent Advances , 2004 .

[29]  Robert N Weinreb,et al.  Comparison of scanning laser polarimetry using variable corneal compensation and retinal nerve fiber layer photography for detection of glaucoma. , 2004, Archives of ophthalmology.

[30]  T. Sejnowski,et al.  Heidelberg retina tomograph measurements of the optic disc and parapapillary retina for detecting glaucoma analyzed by machine learning classifiers. , 2004, Investigative ophthalmology & visual science.

[31]  D Cavouras,et al.  Development of the cubic least squares mapping linear-kernel support vector machine classifier for improving the characterization of breast lesions on ultrasound. , 2004, Computerized medical imaging and graphics : the official journal of the Computerized Medical Imaging Society.

[32]  H. Lemij,et al.  Variable corneal compensation improves discrimination between normal and glaucomatous eyes with the scanning laser polarimeter. , 2004, Ophthalmology.

[33]  D. Greenfield,et al.  Quantitative assessment of structural damage in eyes with localized visual field abnormalities. , 2004, American journal of ophthalmology.

[34]  M. Masotti,et al.  A novel featureless approach to mass detection in digital mammograms based on support vector machines. , 2004, Physics in medicine and biology.

[35]  D. Greenfield,et al.  Retinal Nerve Fiber Layer Assessment Using Scanning Laser Polarimetry , 2004, International ophthalmology clinics.

[36]  F. Medeiros,et al.  Comparison of the GDx VCC scanning laser polarimeter, HRT II confocal scanning laser ophthalmoscope, and stratus OCT optical coherence tomograph for the detection of glaucoma. , 2004, Archives of ophthalmology.

[37]  Hans G Lemij,et al.  The relationship between standard automated perimetry and GDx VCC measurements. , 2004, Investigative ophthalmology & visual science.

[38]  Daniel J. Strauss,et al.  Objective detection of the central auditory processing disorder:A new machine learning approach , 2004, IEEE Transactions on Biomedical Engineering.

[39]  A. Shoeb,et al.  Patient-specific seizure onset detection , 2004, Epilepsy & Behavior.

[40]  Robert N Weinreb,et al.  Confocal scanning laser ophthalmoscopy classifiers and stereophotograph evaluation for prediction of visual field abnormalities in glaucoma-suspect eyes. , 2004, Investigative ophthalmology & visual science.