• Machine algorithms have been developed for automated detection and characterization of glaucomatous defects in visual field records stored in a computer. Using elementary techniques of pattern analysis, the contours defining each visual field are described by a set of primitive graphic and type-descriptive features. A hierarchical structure of logical tests is then used to arrive at intermediate and higher-level conclusions. Decision procedures are tailored for the detection of features commonly found in glaucomatous visual fields. Application of these analysis procedures to a large group of visual field records having a broad range of glaucomatous defects shows a very high sensitivity for detection. Specificity is limited by the extent to which defects are "typical" or obey expected patterns of field loss.
Hart WM. Computer Processing of Visual Data: II. Automated Pattern Analysis of Glaucomatous Visual Fields. Arch Ophthalmol. 1981;99(1):133–136. doi:10.1001/archopht.1981.03930010135019
Artificial Intelligence Resource Center
Customize your JAMA Network experience by selecting one or more topics from the list below.