Pathologic gait diagnosis with computer-averaged electromyographic profiles.

A computer-assisted approach is described which compares EMG profiles from each patient's gait with similar profiles from able-bodied subjects. The EMG and footswitch signals from walking trials were telemetered via a six-channel biotelemetry system; the EMGs were further processed to obtain a linear envelope before A/D conversion into a desktop computer. Simultaneously, a video camera recorded the patient's walking pattern. From 32sec of converted data, a number of strides were selected for averaging to achieve a mean ensemble pattern over the stride period, which is set to 100% for each selected stride. The ensemble average of each muscle's EMG was superimposed on a plot of the EMG patterns from able-bodied subjects. The diagnostician then correlated atypical patterns from each muscle with the stopped or slow motion television image of each patient's gait. Positive corroborative evidence yields detailed diagnostic statements about the cause of their abnormal gait. Such evidence is valuable in planning future rehabilitative procedures or in assessing the results of past efforts.