Abstract
The Computer Assisted Touch Screen (CATS) and Computer Assisted Moving Eye Campimeter (CAMEC) are personal computer (PC)-based video-campimeters which employ multiple and single static stimuli on a cathode ray tube respectively. Clinical studies show that CATS and CAMEC provide comparable results to more expensive conventional visual field test devices. A neural network has been designed to classify visual field data from PC-based video-campimeters to facilitate diagnostic interpretation of visual field test results by non-experts. A three-layer back propagation network was designed, with 110 units in the input layer (each unit corresponding to a test point on the visual field test grid), a hidden layer of 40 processing units, and an output layer of 27 units (each one corresponding to a particular type of visual field pattern). The network was trained by a training set of 540 simulated visual field test result patterns, including normal, glaucomatous and neuro-ophthalmic defects, for up to 20 000 cycles. The classification accuracy of the network was initially measured with a previously unseen test set of 135 simulated fields and further tested with a genuine test result set of 100 neurological and 200 glaucomatous fields. A classification accuracy of 91–97% with simulated field results and 65–100% with genuine field results were achieved. This suggests that neural networks incorporated into PC-based video-campimeters may enable correct interpretation of results in non-specialist clinics or in the community.
Similar content being viewed by others
Article PDF
References
Mutlukan E, Keating D, Damato BE . A touch-screen multi-stimulus video-campimeter. In: Mills R (ed) Perimetry update 92/93. New York: Kugler-Ghedini, 1993:589–95.
Johnston SC, Damato BE, Evans AL, Allan D . Computerised visual field test for children using multiple moving fixation targets. Med Bio-Eng Comput 1989; 27: 612–6.
Mutlukan E . Computerised campimetry with static dark-on-bright stimuli. Doc Ophthalmol 1993; 84: 335–50.
Keating D, Mutlukan E, Damato BE, Evans A . A back propagation neural network for the classification of visual field data. Invest Ophthalmol Vis Sci 1992; 33: 970.
Keating D, Mutlukan E, Evans A, McGarvie J, Damato BE . A back propagation neural network for the classification of visual field data. Phys Med Biol 1993; 38: 1263–70.
Rumelhart DE, McCleland JL . PDP Research Group: parallel distributed processing. In: Explorations in the micro-structure of cognition. Vol. 1. Cambridge, Mass.: MIT Press, 1986.
Mutlukan E, Damato BE . Computerised perimetry with moving and steady fixation in children. Eye 1993; 7: 554–61.
Bajandas FJ, Lanning BJ . Neuro-ophthalmology review manual. Thorofare, New Jersey: Slack Inc., 1988.
Walsh TJ (ed). Visual fields: examination and interpretation. San Francisco: American Academy of Ophthalmology, 1990.
Harrington DO, Drake MV . The visual fields: text and atlas of clinical perimetry. St Louis, Mosby, 1990.
Beck RW, Smith CH . Neuro-ophthalmology: a problem-oriented approach. Boston: Little, Brown, 1988.
Nagata S, Kani K, Sugiyama A . A computer assisted visual field diagnosis system using a neural network. In: Mills RP, Heijl A (eds) Perimetry update 91/92. New York, Kugler-Ghedini, 1991.
Keating D, Mutlukan E, Damato BE, Kirkness CM . Recognition of glaucomatous and neurological field loss with a neural network. Invest Ophthalmol Vis Sci 1993; (Suppl)34: 1417.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Mutlukan, E., Keating, D. Visual field interpretation with a personal computer based neural network. Eye 8, 321–323 (1994). https://doi.org/10.1038/eye.1994.65
Issue Date:
DOI: https://doi.org/10.1038/eye.1994.65
Keywords
This article is cited by
-
Artificial intelligence in diabetic retinopathy
Eye (2021)
-
Pattern recognition can detect subtle field defects in eyes of HIV individuals without retinitis under HAART
Graefe's Archive for Clinical and Experimental Ophthalmology (2011)