Physikalische Medizin, Rehabilitationsmedizin, Kurortmedizin 2009; 19 - A6
DOI: 10.1055/s-0029-1238180

Application of Time-Frequency Surface Electromyographic Procedures for the Classification of Impaired Muscle Function in Musculoskeletal Pain Conditions

G Ebenbichler 1, P Mair 2, C Wirth 3, P Grünbarth 3, P Boissy 4, P Bonato 5
  • 1Physical Medicine & Rehabilitation, Vienna Medical University, Vienna, Austria
  • 22nd Dept of Statistics and Mathematics, Economic University, Vienna, Austria
  • 3Technikum Vienna, Vienna, Austria
  • 4Kinesiology, Université de Sherbrooke & Research Centre on Aging CSSS-IUGS Sherbrooke, Quebec, QC, Canada
  • 5Physical Medicine & Rehabilitation, Harvard Medical School and Harvard-MIT Division of Health Sciences and Technology, Boston, MA, USA

Objective: This pilot study tested whether time-frequency EMG measures allow one to classify impaired back muscle function in chronic back pain (cLBP) as different from pain free healthy subjects.

Design: Case control study.

Setting: Neuromuscular research laboratory, NeuroMuscular Research Center, Boston University (Boston U.S.A.).

Participants: A total of 8 healthy subjects and 8 cLBP patients.

Interventions: Cyclically lifting and lowering of a box for 4 minutes (pace: 12 lifts/minute. The box-load was 13kg in subjects and 8.5kg in patients. EMG signals were recorded from 12 muscle sites from trunk and limb muscles.

Main Outcome Measures: After identification of the lifting and lowering portions of the lifting trials (cycle by cycle), time-frequency analysis methods were applied to compute the Instantaneous Median Frequency (IMDF) for the respective EMG portions of each cycle. Linear regression analysis served to calculate the EMG onsets and changes during the exercise („fatigue slope“).

Results: Patients' mean Roland Morris ratings (mean age: 39.5±12 years) were 10.3±7. At the onset of lifting, median pain ratings on a visual analogue scale were 2.0 (0.25; 3.75). At the end of lifting pain was rated 6.0 (5.25; 8.8). Full EMG data analysis was available from 6 patients and 6 controls. IMDF onset values recorded from the L5 and Th10 recording sites revealed lower values in patients than in controls for the lifting portion of the task cycles. Such difference was not observed for the lowering portion of the task. The fatigue slopes calculated for both the concentric and eccentric portion separately, revealed no relevant differences between patients and healthy subjects at the gluteal and paravertebral back muscle recording sites, although patients lifted 35% less weight.

Conclusions: The IMDF technique has a good potential for augmenting current functional capacity evaluation procedures for patients with musculoskeletal conditions like chronic LBP and may provide a future diagnostic tool for classifying impaired muscle function in these patients.