Volume 47 Number 2, 2010
Pages 133 — 142
Abstract — The entropy of electromyography (EMG) signals suggests a possible tool for the clinical assessment of low back pain (LBP). However, a general physiological mechanism for entropy and pain and/or dysfunction following exercise intervention is still unknown. This comparative study investigated the differences between entropy levels of the EMG signals and the slope of median frequency (MF) based on power spectrum analysis in chronic LBP subjects after 4 weeks of spinal stabilization exercises (SSEs). In total, 46 subjects (24 female and 22 male) participated in the exercise program; the EMG signals of their thoracic and lumbar erector spinae muscles were measured. Following the SSEs, the level of pain as measured by the Million Visual Analog Scale decreased significantly from 3.80 to 2.81 (T = 3.42, p = 0.001). The Shannon entropy levels of the EMG signals were not different based on the back muscles (F = 2.86, p = 0.09) but demonstrated a significant interaction with pain level (F = 7.17, p = 0.01). The slope of MF differed among the muscles (F = 12.06, p = 0.01); however, no interaction with the level of pain was found (F = 0.19, p = 0.66). Therefore, the Shannon entropy of the EMG signals might be a useful and valuable clinical tool to measure pain following intervention.
Key words: complexity, EMG , entropy, erector spinae, low back pain, Million Visual Analog Scale, power spectrum, rehabilitation, slope of median frequency, spinal stabilization exercises.