Trained Artificial Neural Network for Glaucoma Diagnosis Using Visual Field Data: A Comparison With Conventional Algorithms

PurposeTo evaluate and confirm the performance of an artificial neural network (ANN) trained to recognize glaucomatous visual field defects, and compare its diagnostic accuracy with that of other algorithms proposed for the detection of visual field loss. MethodsSITA Standard 30-2 visual fields, from 100 glaucoma patients and 116 healthy participants, formed the data set. Our ANN was a previously described fully trained network using scored pattern deviation probability maps as input data. Its diagnostic accuracy was compared to that of the Glaucoma Hemifield Test, the Pattern Standard Deviation index at the P<5% and <1%, and also to a technique based on the recognizing clusters of significantly depressed test points. ResultsThe included tests had early to moderate visual field loss (median MD=−6.16 dB). ANN achieved a sensitivity of 93% at a specificity level of 94% with an area under the receiver operating characteristic curve of 0.984. Glaucoma Hemifield Test attained a sensitivity of 92% at 91% specificity. Pattern Standard Deviation, with a cut off level at P<5% had a sensitivity of 89% with a specificity of 93%, whereas at P<1% the sensitivity and specificity was 72% and 97%, respectively. The cluster algorithm yielded a sensitivity of 95% and a specificity of 82%. ConclusionsThe high diagnostic performance of our ANN based on refined input visual field data was confirmed in this independent sample. Its diagnostic accuracy was slightly to considerably better than that of the compared algorithms. The results indicate the large potential for ANN as an important clinical glaucoma diagnostic tool.

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

[2]  A Heijl,et al.  Glaucoma Hemifield Test. Automated visual field evaluation. , 1992, Archives of ophthalmology.

[3]  G. Lindgren,et al.  The effect of perimetric experience in normal subjects. , 1989, Archives of ophthalmology.

[4]  B. Bengtsson,et al.  The effect of perimetric experience in patients with glaucoma. , 1996, Archives of ophthalmology.

[5]  B. Bengtsson,et al.  Inter-subject variability and normal limits of the SITA Standard, SITA Fast, and the Humphrey Full Threshold computerized perimetry strategies, SITA STATPAC. , 1999, Acta ophthalmologica Scandinavica.

[6]  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.

[7]  J. Hanley,et al.  The meaning and use of the area under a receiver operating characteristic (ROC) curve. , 1982, Radiology.

[8]  C. Bunce,et al.  The ability of the GDx Nerve Fibre Analyser neural network to diagnose glaucoma , 2001, Graefe's Archive for Clinical and Experimental Ophthalmology.

[9]  Xiaohui Liu,et al.  A spatio-temporal Bayesian network classifier for understanding visual field deterioration , 2005, Artif. Intell. Medicine.

[10]  Anders Heijl,et al.  Effects of input data on the performance of a neural network in distinguishing normal and glaucomatous visual fields. , 2005, Investigative ophthalmology & visual science.

[11]  Georg Lindgren,et al.  A package for the statistical analysis of visual fields , 1987 .

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

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

[14]  T. Sejnowski,et al.  Unsupervised machine learning with independent component analysis to identify areas of progression in glaucomatous visual fields. , 2005, Investigative ophthalmology & visual science.

[15]  A Heijl,et al.  Arcuate cluster analysis in glaucoma perimetry. , 1993, Journal of glaucoma.

[16]  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.

[17]  L N Johnson,et al.  Visual field interpretation with empiric probability maps. , 1989, Archives of ophthalmology.

[18]  Terrence J. Sejnowski,et al.  Comparison of machine learning and traditional classifiers in glaucoma diagnosis , 2002, IEEE Transactions on Biomedical Engineering.

[19]  J. Wild,et al.  The influence of the learning effect on automated perimetry in patients with suspected glaucoma , 1989, Acta ophthalmologica.

[20]  A Heijl,et al.  Spatial analyses of glaucomatous visual fields; a comparison with traditional visual field indices , 1992, Acta ophthalmologica.

[21]  C. Glymour,et al.  Optical coherence tomography machine learning classifiers for glaucoma detection: a preliminary study. , 2005, Investigative ophthalmology & visual science.

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

[23]  Mei-Ling Huang,et al.  Development and comparison of automated classifiers for glaucoma diagnosis using Stratus optical coherence tomography. , 2005, Investigative ophthalmology & visual science.

[24]  Balwantray C. Chauhan,et al.  A cluster analysis for threshold perimetry , 2005, Graefe's Archive for Clinical and Experimental Ophthalmology.

[25]  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.

[26]  J Katz,et al.  Comparison of analytic algorithms for detecting glaucomatous visual field loss. , 1991, Archives of ophthalmology.

[27]  B. Bengtsson,et al.  Early Visual Field Defects in Glaucoma: A Study of Eyes Developing Field Loss , 1996 .

[28]  T. Sejnowski,et al.  Relevance vector machine and support vector machine classifier analysis of scanning laser polarimetry retinal nerve fiber layer measurements. , 2005, Investigative ophthalmology & visual science.

[29]  M. Zweig,et al.  Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. , 1993, Clinical chemistry.

[30]  Robert N Weinreb,et al.  Using unsupervised learning with variational bayesian mixture of factor analysis to identify patterns of glaucomatous visual field defects. , 2004, Investigative ophthalmology & visual science.

[31]  Robert N Weinreb,et al.  Using unsupervised learning with independent component analysis to identify patterns of glaucomatous visual field defects. , 2005, Investigative ophthalmology & visual science.

[32]  Douglas R. Anderson Automated Static Perimetry , 1992 .

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

[34]  Anders Heijl,et al.  Essential perimetry: The field analyzer primer , 2002 .

[35]  A. Heijl,et al.  Evaluation of methods for automated Hemifield analysis in perimetry. , 1992, A M A Archives of Ophthalmology.