advertisement

Editors Selection IGR 7-1

Visual function: Artificial intelligence

Comment by Joseph Caprioli on:

11704 Relevance vector machine and support vector machine classifier analysis of scanning laser polarimetry retinal nerve fiber layer measurements, Bowd C; Medeiros FA; Zhang Z et al., Investigative Ophthalmology and Visual Science, 2005; 46: 1322-1329

Find related abstracts

Artificial intelligence techniques have been increasingly applied to the study of glaucoma. Neural networks and machine classifiers have been used to help classify subjects as glaucomatous or normal on the basis of a number of (sometimes complex) visual function measurements and optic nerve structural measurements, or a combination of both. Bowd et al. (151) use the techniques of two advanced learning classifiers trained with results from scanning laser polarimetric measurements of nerve fiber layer thickness. Normal controls and glaucoma patients with moderate visual field loss were so tested and classified. The results indicate good potential for such classification approaches, particularly when the input data set is complex, and might be intuitively difficult for a clinician to interpret. Even though a cross-validation technique was used, true validation of these results requires testing a completely different group of subjects. Whether similar approaches will help with the identification of really early damage remains to be demonstrated. This reviewer would suggest that such identification would require the combination of data from several structural and perhaps functional approaches.

• ← Previous
• Next →