The Higher-Order Statistics Applied to Electromyographic Signals of Persons With Low Back Pain to Improve the Information Content of a Physiological Examination

Abstract

Low back pain (LBP) is a common global problem as it is the leading cause of years lived with disability. In the case of LBP, electromyography (EMG) is used for diagnostic purposes and monitoring the functional state of a person. Electromyographic signals (EMGs) are usually analyzed using 2nd-order statistical methods, such as power spectrum, but due to the non-stationarity, nonlinearity and non-Gaussianity of EMGs, these methods cannot provide an adequate analysis. Thus, higher-order statistical methods are useful. This work is devoted to the study of higher-order statistical characteristics of EMGs for LBP persons. Signals from five groups of patients, including vertebral disorders, scoliosis, functional pain, healthy people without complaints, and healthy people with pain complaints were processed.