Classification of back muscle impairment based on the surface electromyographic signal.

A surface electromyographic (EMG) procedure for classifying muscle impairments in persons with low back pain (LBP) is described. The procedure was studied using a device, the Back Analysis System (BAS), to acquire and process EMG signals from six bilateral muscle sites during sustained isometric contractions designed to progressively fatigue the lower back. Back muscle impairment was determined on the basis of the different ways in which the EMG median frequency parameters change as a function of contraction duration and muscle site. The article describes a series of studies that have been useful in developing an automated procedure for identifying back muscle impairment by comparing individual test results to a normative database. To date, the research results have produced multivariate discriminant functions that have identified two muscle impairment categories associated with deconditioning and imbalances secondary to LBP. We have found that the functions can distinguish individuals with and without LBP with an accuracy of approximately 90%. Other studies are described in which the technique is applied to monitoring changes in muscle performance capability that occur following rehabilitation for LBP. Many of our findings here are also compared to the results of independent studies by others using similar procedures. The need for further research and development of the technique to improve its clinical applicability is also described.

Development of new protocols and analysis procedures for the assessment of LBP by surface EMG techniques.

Spectral parameters of the surface electromyographic (EMG) signal from lumbar back muscles assessed during a fatiguing isometric contraction can be used to classify different categories of low back pain (LBP) subjects and control subjects without LBP. In the test protocol currently used at the NeuroMuscular Research Center at Boston University, subjects contract their back muscles at 80% of their maximal voluntary contraction (MVC) force. This fatigue-based protocol has been successfully applied to persons with subacute or chronic LBP; those in acute pain, however, have not been included because of their inability to perform a maximal exertion. In this paper we will examine the force sensitivity of the currently used EMG parameters and also give an overview of some of our efforts to develop new test procedures. Our goal is to develop force-insensitive surface EMG parameters that can be used for classification purposes in populations of subjects who develop low trunk extension forces. In addition, the development of a model to predict MVC from anthropometrical measurements will be presented.

Spectral electromyographic assessment of back muscles in patients with low back pain undergoing rehabilitation.

STUDY DESIGN: A surface electromyographic procedure for evaluating back muscle impairment was studied in patients undergoing rehabilitation for low back pain. OBJECTIVES: The results were analyzed to determine whether the electromyographic procedure was able to: 1) distinguish muscle impairment between patients with low back pain and normal subjects, and 2) monitor changes in muscle function after low back pain rehabilitation. METHODS: Patients with chronic low back pain (n = 85) were tested to measure the median frequency of the electromyographic signals from six lumbar electrode sites during sustained trunk extensions. A subset (n = 28) of these patients was re-tested after low back pain rehabilitation. A discriminant function for classifying subjects into "low back pain" and "normal" groups was formulated using the electromyographic data from a subset of the patients with low back pain (n = 28) and a normative sample (n = 42). Results for this "learning" sample were compared with results using the same function on the remaining "holdout" sample of patients (n = 57) and an additional normative sample (n = 6). Differences in electromyographic parameters before and after rehabilitation also were analyzed. RESULTS: The discriminant function classified subjects into low back pain and normal groups, with 86% and 89% correct classification for the "learning" and "holdout" samples, respectively. These classification results were independent of trunk extensor strength. Changes in median frequency after the rehabilitation program were consistent with improvements in back muscle fatigability. CONCLUSION: These findings demonstrate how electromyographic spectral measurements may be used to identify and monitor back muscle impairment in patients undergoing rehabilitation for low back pain.