Situation assessment of glaucoma using a hybrid fuzzy neural network

Describes a differential diagnostic-decision support system to aid in early detection in primary-care environments. In general, the authors are able to give correct differential diagnostic support under primary-care conditions without any further financial investments. In some cases, it is difficult to differentiate between "glaucomatous" and "pathological" situation classes. There is an unclear transition between these two situations. Initial tests of a combination crisp/fuzzy output refined the authors' results, and these are now under investigation. More data are needed for the "normal" class. An evaluation of the complete monitoring system will be started, including all components and sampling of more data sets. For consecutive patient cases, time-dependent characteristic changes in the visual fields could be detected. Further developmental work will be done in this area, parallel to the growth of the database.