Effect of Normalization Methods on the Reliability of EMG during Stair Negotiation and Ramp Walking

Due to the variability of the Electromyography (EMG) signals in different individuals, the signals need to be normalized to a reference signal in order to enable comparisons between individuals. The aim of the present study was to determine the reliability of EMG of lower extremity muscles in five signal normalization methods.

A total of 14 healthy individuals ascended and descended a set of stairs and a ramp. Each task was repeated six times. Surface EMGs of the Vastus Medialis (VM), Vastus Lateralis (VL), Rectus Femuris (RF), and Tibialis Anterior (TA) muscles were recorded during each exercise and the isometric Maximum Voluntary Contraction (MVC) of the muscles. Electromyography signals were normalized using five different methods. The reference signal for the normalization was corresponding signal of each muscle in MVC, Mean Dynamic Task (MT), Peak Dynamic Task Activity (PT), Mean of Raw Signal (MT-raw), and Peak of Raw Signal (PT-raw).

The MVC method showed the highest reliability and the lowest standard measurement error (SEM). The averages obtained for ICC and SEM for the MVC method in different activities were 0.74 and 4.8, respectively. Also, the MT method showed the lowest reliability and the highest error. The averages obtained for ICC and SEM in different exercises were found to be 0.15 and 88.43, respectively. Although PT, PT-raw, and MT-raw methods showed a similar and average reliability (0.33

In conclusion, MVC demonstrated the highest reliability; therefore, it is suggested as the best normalization method for young healthy people during stair up and ramp walking.