Serge H. Roy

Filtering the surface EMG signal: Movement artifact and baseline noise contamination

The surface electromyographic (sEMG) signal that originates in the muscle is inevitably contaminated by various noise signals or artifacts that originate at the skin-electrode interface, in the electronics that amplifies the signals, and in external sources. Modern technology is substantially immune to some of these noises, but not to the baseline noise and the movement artifact noise. These noise sources have frequency spectra that contaminate the low-frequency part of the sEMG frequency spectrum. There are many factors which must be taken into consideration when determining the appropriate filter specifications to remove these artifacts; they include the muscle tested and type of contraction, the sensor configuration, and specific noise source. The band-pass determination is always a compromise between (a) reducing noise and artifact contamination, and (b) preserving the desired information from the sEMG signal. This study was designed to investigate the effects of mechanical perturbations and noise that are typically encountered during sEMG recordings in clinical and related applications. The analysis established the relationship between the attenuation rates of the movement artifact and the sEMG signal as a function of the filter band pass. When this relationship is combined with other considerations related to the informational content of the signal, the signal distortion of filters, and the kinds of artifacts evaluated in this study, a Butterworth filter with a corner frequency of 20 Hz and a slope of 12 dB/oct is recommended for general use. The results of this study are relevant to biomechanical and clinical applications where the measurements of body dynamics and kinematics may include artifact sources.

Median frequency of the myoelectric signal: effects of muscle ischemia and cooling

SummaryA study was performed to investigate the changes that occur in the median frequency of the myoelectric signal during local ischemia or reduction of intramuscular temperature produced by surface cooling. Data was obtained from experiments which involved the first dorsal interosseous muscle of 10 female and 16 male subjects. These subjects were asked to perform isometric constant-force abduction contractions of the index finger at 20% and 80% of maximal voluntary contraction level. The initial median frequency (IMF) of the myoelectric signal during the first 0.5 s of contraction was calculated. Results showed a significant reduction of the IMF in contractions performed under ischemic conditions; upon release, the IMF recovered quickly. At 80% maximal voluntary level of contraction, a greater decrease of the IMF was recorded. Similar results were demonstrated during reduction of intramuscular temperature with gradual recovery of the IMF after cooling. These results demonstrate that the median frequency of the myoelectric signal displays behavior similar to that reported for conduction velocity and this is consistent with the notion that accumulation of metabolic byproducts in muscle tissue causes a decrease in the conduction velocity of the muscle fibers.

Fatigue, recovery, and low back pain in varsity rowers

The purpose of this study was to determine whether surface electromyography (EMG) from the erector spinae muscles could correctly identify individuals with low back pain without a population of elite athletes. A similar technique had previously been successful in identifying low back pain patients within a non-athletic population. A Back Analysis System was used to compute the median frequency of the EMG power density spectrum to monitor metabolic changes in back muscles associated with muscle fatigue. Twenty-three members of a mens collegiate varsity crew team consisting of port (N = 13) and starboard (N = 10) rowers were tested in a laboratory during a fatigue-inducing isometric contraction sustained at a relatively high, constant force. Six of the rowers tested were further classified as having low back pain. A brief test contraction was repeated at a fixed interval following the fatiguing contraction to monitor recovery. A two-group discriminant analysis procedure correctly classified 100% of the rowers with low back pain and 93% of the rowers without back pain on the basis of the median frequency data. The median frequency parameters related to recovery were the best discriminators of back pain. A similar analysis correctly classified 100% of the port rowers and 100% of the starboard rowers on the basis of their spectral parameters. The best discriminating variables in this instance were the median frequency parameters relating to both fatigability and recovery. Results from this study demonstrate that low back pain and asymmetrical muscle function in rowers can be assessed on the basis of EMG spectral analysis.

Inter-electrode spacing of surface EMG sensors: Reduction of crosstalk contamination during voluntary contractions

We investigated the influence of inter-electrode spacing on the degree of crosstalk contamination in surface electromyographic (sEMG) signals in the tibialis anterior (target muscle), generated by the triceps surae (crosstalk muscle), using bar and disk electrode arrays. The degree of crosstalk contamination was assessed for voluntary constant-force isometric contractions and for dynamic contractions during walking. Single-differential signals were acquired with inter-electrode spacing ranging from 5 mm to 40 mm. Additionally, double differential signals were acquired at 10 mm spacing using the bar electrode array. Crosstalk contamination at the target muscle was expressed as the ratio of the detected crosstalk signal to that of the target muscle signal. The crosstalk contamination ratio approached a mean of 50% for the 40 mm spacing for triceps surae muscle contractions at 80% MVC and tibialis anterior muscle contractions at 10% MVC. For single differential recordings, the minimum crosstalk contamination was obtained from the 10 mm spacing. The results showed no significant differences between the bar and disk electrode arrays. During walking, the crosstalk contamination on the tibialis anterior muscle reached levels of 23% for a commonly used 22 mm spacing single-differential disk sensor, 17% for a 10 mm spacing single-differential bar sensor, and 8% for a 10 mm double-differential bar sensor. For both studies the effect of electrode spacing on crosstalk contamination was statistically significant. Crosstalk contamination and inter-electrode spacing should therefore be a serious concern in gait studies when the sEMG signal is collected with single differential sensors. The contamination can distort the target muscle signal and mislead the interpretation of its activation timing and force magnitude.

Detecting freezing-of-gait during unscripted and unconstrained activity

We present a dynamic neural network (DNN) solution for detecting instances of freezing-of-gait (FoG) in Parkinsons disease (PD) patients while they perform unconstrained and unscripted activities. The input features to the DNN are derived from the outputs of three triaxial accelerometer (ACC) sensors and one surface electromyographic (EMG) sensor worn by the PD patient. The ACC sensors are placed on the shin and thigh of one leg and on one of the forearms while the EMG sensor is placed on the shin. Our FoG solution is architecturally distinct from the DNN solutions we have previously designed for detecting dyskinesia or tremor. However, all our DNN solutions utilize the same set of input features from each EMG or ACC sensor worn by the patient. When tested on experimental data from PD patients performing unconstrained and unscripted activities, our FoG detector exhibited 83% sensitivity and 97% specificity on a per-second basis.

EMG-based measures of fatigue during a repetitive squat exercise

We have demonstrated a technique to calculate the EMG instantaneous median frequency to assess muscle fatigue during a dynamic exercise commonly prescribed in patients with ACL deficiency. We used Cohen-Posch time-frequency representations to improve upon the variability of the instantaneous median frequency estimates derived using Cohen Class transformations. The technique was applied to surface EMG data recorded from the quadriceps and hamstring muscles of a control subject and a patient with ACL deficiency during a repetitive squat exercise. Instantaneous median frequency values were derived for the knee-extension phases of the exercise. Ensemble average and standard deviation of the instantaneous median frequency were computed for the portion of the cycle associated with the lowest variability of the mechanics.

High-Resolution Tracking of Motor Disorders in Parkinson's Disease During Unconstrained Activity

Parkinsons disease (PD) can present with a variety of motor disorders that fluctuate throughout the day, making assessment a challenging task. Paper‐based measurement tools can be burdensome to the patient and clinician and lack the temporal resolution needed to accurately and objectively track changes in motor symptom severity throughout the day. Wearable sensor‐based systems that continuously monitor PD motor disorders may help to solve this problem, although critical shortcomings persist in identifying multiple disorders at high temporal resolution during unconstrained activity. The purpose of this study was to advance the current state of the art by (1) introducing hybrid sensor technology to concurrently acquire surface electromyographic (sEMG) and accelerometer data during unconstrained activity and (2) analyzing the data using dynamic neural network algorithms to capture the evolving temporal characteristics of the sensor data and improve motor disorder recognition of tremor and dyskinesia. Algorithms were trained (n = 11 patients) and tested (n = 8 patients; n = 4 controls) to recognize tremor and dyskinesia at 1‐second resolution based on sensor data features and expert annotation of video recording during 4‐hour monitoring periods of unconstrained daily activity. The algorithms were able to make accurate distinctions between tremor, dyskinesia, and normal movement despite the presence of diverse voluntary activity. Motor disorder severity classifications averaged 94.9% sensitivity and 97.1% specificity based on 1 sensor per symptomatic limb. These initial findings indicate that new sensor technology and software algorithms can be effective in enhancing wearable sensor‐based system performance for monitoring PD motor disorders during unconstrained activities.

Preferred sensor sites for surface EMG signal decomposition

Technologies for decomposing the electromyographic (EMG) signal into its constituent motor unit action potential trains have become more practical by the advent of a non-invasive methodology using surface EMG (sEMG) sensors placed on the skin above the muscle of interest (De Luca et al 2006 J. Neurophysiol. 96 1646-57 and Nawab et al 2010 Clin. Neurophysiol. 121 1602-15). This advancement has widespread appeal among researchers and clinicians because of the ease of use, reduced risk of infection, and the greater number of motor unit action potential trains obtained compared to needle sensor techniques. In this study we investigated the influence of the sensor site on the number of identified motor unit action potential trains in six lower limb muscles and one upper limb muscle with the intent of locating preferred sensor sites that provided the greatest number of decomposed motor unit action potential trains, or motor unit yield. Sensor sites rendered varying motor unit yields throughout the surface of a muscle. The preferred sites were located between the center and the tendinous areas of the muscle. The motor unit yield was positively correlated with the signal-to-noise ratio of the detected sEMG. The signal-to-noise ratio was inversely related to the thickness of the tissue between the sensor and the muscle fibers. A signal-to-noise ratio of 3 was found to be the minimum required to obtain a reliable motor unit yield.

Clinician's view: dynamic EMG

The research into a correlation between joint biomechanics and the action of muscles that act on the limb segment involved during movement has become very important in recent years. However, while the techniques for elaborating the EMG signal and its relationship with the dynamics of movement are described in detail in the literature, from a clinical point of view publications on how these techniques are used for clinical gait analysis applications are scarce. The purpose of dynamic EMG in clinical gait analysis is essentially to define the muscular activity that controls joint movement during gait, as shown by studies carried out on children with cerebral palsy, in which the abnormal pattern of muscle activation is used, for example, as an indication for surgical tendon transfer or lengthening.

Dynamic neural network detection of tremor and dyskinesia from wearable sensor data

We present a dynamic neural network (DNN) solution for detecting time-varying occurrences of tremor and dyskinesia at 1 s resolution from time series data acquired from surface electromyographic (sEMG) sensors and tri-axial accelerometers worn by patients with Parkinsons disease (PD). The networks were trained and tested on separate datasets, each containing approximately equal proportions of tremor, dyskinesia, and disorder-free data from 8 PD and 4 control subjects performing unscripted and unconstrained activities in an apartment-like environment. During DNN testing, tremor was detected with a sensitivity of 93% and a specificity of 95%, while dyskinesia was detected with a sensitivity of 91% and a specificity of 93%. Similar sensitivity and specificity levels were obtained when DNN testing was carried out on subjects who were not included in DNN training.