Computer-assisted assessment of cochlear nerve fibers.

Quantitative assessment of cochlear nerve fibers, with statistical analysis of histologic specimens, is facilitated by the application of computer image processing techniques. Three algorithms, developed for the evaluation of cochlear sections, are presented, MANUAL, AIDED, and AUTO, in order of increasing degree of automation and decreasing requirement for operator intervention. All three algorithms demonstrate high accuracy and consistency in identifying cochlear nerve fibers and quickly provide statistics regarding the target nerve fiber population. These algorithms have potential utility in the study of cochlear nerve pathologic alterations, as for example, with intracochlear electrical stimulation.