Visually displayed EMG feedback: single case studies of hemiplegic upper extremity rehabilitation

The efficacy of visually displayed EMG feedback in treating hemiplegic upper limb motor disorders was investigated in 5 patients (0.5-4 years poststroke). A single case experimental method "multiple-baseline across target behaviors" was used to compare performance during the feedback phase to that occurring in the monitored baseline phase. The nonfeedback baseline phase was followed by the staggered introduction of audiovisual feedback for each of the targeted pairs of muscles. EMG feedback obtained from muscle pairs (shoulder: anterior deltoid and upper trapezius; elbow: brachial triceps and biceps; digits: extensor digitorum communis and digit flexors) was displayed as two distinct waveforms on a videomonitor during therapy. Nonfeedback assessment of each of the three pairs was performed during each session. The effects of feedback were not uniformly distributed. Elbow control responded best, and statistical tests confirmed the effects of intervention observed clinically in all 5 patients. Clinical improvements in shoulder flexion were observed in 4 patients but could be statistically attributed to EMG gains in just 1. Improvement in finger extension observed clinically in 3 patients was statistically attributable to EMG gains in 2. All patients regained control of at least one target activity. EMG gains were accompanied by increases in active range of motion and by varying functional improvement. Marked functional gains in 3 patients were obtained with recovery of finger control.