engleski

Surface EMG Frequency-Based Discrimination of Low Back Pain Patients

The back muscle endurance is clearly identified as a predictor of the first occurrence and recurrence of low back pain, with considerable amount of research being performed on this topic. The frequency parameters-based evaluation of the issued low back pain (LBP) pathology was deployed, utilizing the Hilbert-Huang (HHT) and Fourier (STFT) analysis techniques for the induced lumbar muscle fatigue assessment by means of surface electromyography (sEMG). The time-dependent median frequency estimate is used as muscle fatigue indicator, and linear regression parameters, slope (k) and initial frequency (f0), are derived as fatigue quantifiers. A total number of 92 male subjects were considered, divided in three groups: 46 healthy subjects (HS), 28 subjects with chronic low back pain (CLBP) and the remaining 18 subjects with radicular low back pain (RLBP). Four channel sEMG recordings for each subject were acquired where upper (ULES) and lower (LLES) lumbar erector spinae muscles, both in left-right pairs, were tracked. By employing the t-test statistics (α=0.05), significant relative difference on slope k, in left-right muscle relation, for RLBP toward HS and CLPB groups is reported (p<0.05), as well as no significant difference between HS and CLBP (p>0.05), both for ULES and LLES muscles separately. Also, the slope k ratio between ULES and LLES muscles shows the significant difference for CLBP and RLBP groups toward HS (p<0.05), where no significant difference between CLBP and RLBP is established (p>0.05). In this initial research, it is shown that the regression slope k parameter, presenting the time-rate of fatiguing process and also the muscle involvement in a particular contraction, is a very good indicator for discriminating our groups, with HHT providing statistically more significant results then the STFT based analysis. The initial frequency parameter (f0) does not show the statistically significant results to be considered as a reliable LBP pathology indicator.

ow back pain ; time-frequency analysis ; Hilbert-Huang transform ; surface electromyography ; sEMG ; muscle fatigue

nije evidentirano