The goal of this project is to develop a software product which can indicate the onset of carpal tunnel syndrome (CTS) in computer keyboard operators. This software will apply pattern analysis methods to keystroke time series data acquired during normal typing at a microcomputer keyboard. The specific Phase I objective is to perform a pilot study to identify patterns in the keystroke data which might correlate well with early stages of CTS. We have identified a number of analysis methods which we will apply to keystroke data acquired during typing tests given to a subject sample consisting of typists with and without CTS. The results of the analysis will be a set of metrics. We will evaluate these metrics on the basis of variance estimates and receiver operating characteristic (ROC) analysis. The ROC analysis will enable us to determine which methods hold greatest promise in terms of diagnostic accuracy with regard to CTS. The Phase I results will provide us with the analytical basis for performing prospective studies of early CTS detection in Phase II. The resulting software product will be available as a diagnostic tool which can be operated on standard microcomputers with no additional hardware requirements.
