Development and application of a protocol for fine-wire and surface EMG Data collection as part of clinical gait assessment

Onmanee, P 2016, Development and application of a protocol for fine-wire and surface EMG Data collection as part of clinical gait assessment , PhD thesis, University of salford.

PDF (PhD thesis)
Download (5MB) | Preview


Background: Electromyography (EMG) is a measure of neural activation to muscles and as such can give a window into neuromuscular dysfunction in patients. Although it was the primary focus of early clinical gait analysis (CGA), it has become progressively less common since the widespread adoption of optoelectronic measuring systems capturing three dimensional kinematics and kinetics. This is surprising since EMG has considerable potential to explain gait deviations observed in the kinematic and kinetic data. Apart from the extra time required for collecting data there are a number of barriers to the use of EMG in modern CGA. The most obvious is that EMG data has traditionally been collected, analysed and, most importantly, presented using quite different techniques which prevents a streamlined integration of EMG with the kinematic and kinetic data. Secondly, although the general characteristics of normative EMG patterns in the larger muscles are reasonably well understood, there is considerably less consensus on those which are smaller but still clinically important. Finally several of the most clinically important muscles, such as the tibialis posterior (TP), can only be accessed using fine-wire techniques and there is no consensus on how such data should be presented clinically. Objectives: This research aims to define rigorous data capture, analysis and presentation protocols for incorporation of both fine-wire and surface EMG measurements into CGA. The secondary aim is to provide definitive normative EMG profiles in the selected lower limb muscles across the gait cycle in healthy adults as reference for CGA purposes. Finally, a case series aim to explore whether the methods of collection, analysis and data presentation established in this work could be used to detect patterns of muscle dysfunction underlying kinematic impairments in the gait of stroke participants. Methods/results/discussion: A systematic review was conducted and the synthesised EMG profiles with and without between-subject variability from all included papers showed a wide range of variability in lower limb EMG profiles, a lack of studies in deep muscles which potentially play important roles in gait such as TP, no standardisation of fine-wire EMG acquisition and processing (compared to the surface EMG) and various methods of EMG normalisation. These variety of collection and analysis techniques resulted in large variability, in the current literature base, of EMG profiles between different studies. The majority of EMG studies currently available in the literature focus on larger superficial muscles. Studies on TP were scarce in spite of its important role in foot posture and gait. One reason for the lack of information on deep lower limb muscles may be that these can only be assessed using fine-wire sensors, for which there are no guidelines for standardised collection procedures amenable for use in CGA. A series of experiments aimed at addressing these limitations of fine-wire EMG in the current literature base (identified in the systematic review) and ultimately using improved collection and analyses techniques to allow direct comparison between fine-wire and surface EMG and provide a normative database for clinical application were carried out on TP for which little normative reference data exists, tibialis anterior (TA), and medial gastrocnemius (MG). The normalisation study mean normalisation appears to be the best method to reduce variability and this is true across muscles, sensors and different measures of variability: standard deviation can be reduced by 18%-62% of the mean signal and standard errors of measurement can be reduced up to 42% of the mean signal. A peak normalisation is equally effective with small difference (<5%). The second study revealed six gait cycles were necessary to collect fine-wire EMG which showed similar patterns ( r >0.9) at the same standard error of an ensemble average of surface EMG for TA and MG. The grand ensemble average of fine-wire EMG showed slightly greater between-session variability than surface EMG (9%-10% for fine-wire and 4%-7% for surface). Normative EMG data was then collected using normalisation with respect to the peak over six gait cycles from TP, TA and MG alongside kinematics and kinetics at five different speeds from eight young participants. Finally a case series of EMG collections with participants with stroke were used to explore the proof-of-concept of how standardised EMG methods could be implemented in clinical gait analysis and the potential benefits of using EMG to support identification of reasons for gait deviations in CGA. A normative database collected using these established methods was effective to identify pathological features and changes of muscle activity in three participants with post-stroke when using ankle-foot orthosis (AFO). However, the sensitivity of the database to detect changes under AFO condition depended on the severity of the impairment. Conclusion: As there was no previous standardised guidelines for the use of fine-wire EMG in CGA, this PhD defined a protocol for EMG measurement of TP, TA and MG using fine-wire and surface sensors in combination with kinetics and kinematics for CGA. The results of a series of systematic examinations of different normalisation techniques as well as between subject and between-session variability indicate that six gait cycles of data is sufficient for the collection of fine-wire EMG in CGA and that normalisation relative to the mean or peak during the gait cycle is the most appropriate if EMG data is to be used to aide CGA. A case series of stroke participants demonstrated data collected in this way could be used to detect impaired muscle activation underlying impaired kinematics of walking when compared to a normative database, and that the EMG data could add useful information to understanding typical CGA outputs.

Item Type: Thesis (PhD)
Schools: Schools > School of Health Sciences
Funders: University of Salford
Depositing User: P Onmanee
Date Deposited: 08 Dec 2016 09:43
Last Modified: 08 Dec 2016 09:43

Actions (login required)

Edit record (repository staff only) Edit record (repository staff only)


Downloads per month over past year