Computer-based Classification of Visual Field Data
Purpose: Since many neuroophthalmological diseases and lesions may be recognized from perimetric examinations, the appropriate classification of visual field data is essential for diagnosis. However, adequate classification and interpretation of perimetric examination results is not a trivial task and requires well-trained personnel with long-term experience. Therefore, an Internet-based classification system for visual field data derived from perimetric examinations is introduced that may act as a world-wide accessible "counselor" to the diagnosing physician.
Methods: The classification system (see below) is based on a neural Hopfield-attractor-network that consists of N neurons. These neurons are assigned to the N stimulus locations of the stimulus grid that is used to examine the visual field (temporary restriction: only TAP-2000 stimulus grid for a 30°-visual field available). Therefore, the neurons obtain their input data from perimetric examination results being submitted through the Internet (). An iterated relaxation process determines the states of the neurons dynamically. Therefore, even "noisy" perimetric output, e. g., early stages of a disease, may eventually be classified correctly according to the predefined idealized scotomata patterns (diseases) stored as attractors of the network. The classification result containing confidence levels for each network-known disease is transmitted via e-mail or through the Internet to the sender almost immediately after data-submission.
Results: Preliminary tests of the classification system on real visual field data derived from perimetric examinations have shown a classification success of over 80%. There is good reason to believe that this success rate might be improved in future versions of the classification system.
Conclusions: An Internet-based classification system for visual field data, like the one proposed here, furnishes a valuable first overview in judging perimetric examination results and gives an indication towards the final diagnosis by a physician. It should not be considered a substitute for the diagnosing physician. However, thanks to the world-wide accessibility of the Internet the classification system offers a promising perspective towards modern computer-assisted diagnosis in medicine.
Wolfgang Fink, Ph. D.
Senior Researcher at Jet Propulsion Laboratory (JPL)
Visiting Research Associate Professor of Ophthalmology at University of Southern California (USC)
Visiting Research Associate Professor of Neurosurgery at University of Southern California (USC)
Visiting Associate in Physics at California Institute of Technology (Caltech)
California Institute of Technology
Visual and Autonomous Exploration Systems Research Laboratory
15 Keith Spalding (corner of E. California Blvd & S. Wilson Ave)
Mail Code 103-33
Pasadena, CA 91125