Verity M. McClelland

Rectification of the EMG is an unnecessary and inappropriate step in the calculation of Corticomuscular coherence

Corticomuscular coherence (CMC) estimation is a frequency domain method used to detect a linear coupling between rhythmic activity recorded from sensorimotor cortex (EEG or MEG) and the electromyogram (EMG) of active muscles. In motor neuroscience, rectification of the surface EMG is a common pre-processing step prior to calculating CMC, intended to maximize information about action potential timing, whilst suppressing information relating to motor unit action potential (MUAP) shape. Rectification is believed to produce a general shift in the EMG spectrum towards lower frequencies, including those around the mean motor unit discharge rate. However, there are no published data to support the claim that EMG rectification enhances the detection of CMC. Furthermore, performing coherence analysis after the non-linear procedure of rectification, which results in a significant distortion of the EMG spectrum, is considered fundamentally flawed in engineering and digital signal processing. We calculated CMC between sensorimotor cortex EEG and EMG of two hand muscles during a key grip task in 14 healthy subjects. CMC calculated using unrectified and rectified EMG was compared. The use of rectified EMG did not enhance the detection of CMC, nor was there any evidence that MUAP shape information had an adverse effect on the CMC estimation. EMG rectification had inconsistent effects on the power and coherence spectra and obscured the detection of CMC in some cases. We also provide a comprehensive theoretical analysis, which, along with our empirical data, demonstrates that rectification is neither necessary nor appropriate in the calculation of CMC.

Central motor conduction studies and diagnostic magnetic resonance imaging in children with severe primary and secondary dystonia

Aim  Dystonia in childhood has many causes. Imaging may suggest corticospinal tract dysfunction with or without coexistent basal ganglia damage. There are very few published neurophysiological studies on children with dystonia; one previous study has focused on primary dystonia. We investigated central motor conduction in 62 children (34 males, 28 females; age range 3–19y, mean age 10y 8mo, SD 4y 8mo) with severe dystonia to evaluate corticospinal tract integrity before consideration for deep brain stimulation.

Differences in globus pallidus neuronal firing rates and patterns relate to different disease biology in children with dystonia

Background The pathophysiology underlying different types of dystonia is not yet understood. We report microelectrode data from the globus pallidus interna (GPi) and globus pallidus externa (GPe) in children undergoing deep brain stimulation (DBS) for dystonia and investigate whether GPi and GPe firing rates differ between dystonia types. Methods Single pass microelectrode data were obtained to guide electrode position in 44 children (3.3–18.1 years, median 10.7) with the following dystonia types: 14 primary, 22 secondary Static and 8 progressive secondary to neuronal brain iron accumulation (NBIA). Preoperative stereotactic MRI determined coordinates for the GPi target. Digitised spike trains were analysed offline, blind to clinical data. Electrode placement was confirmed by a postoperative stereotactic CT scan. Findings We identified 263 GPi and 87 GPe cells. Both GPi and GPe firing frequencies differed significantly with dystonia aetiology. The median GPi firing frequency was higher in the primary group than in the secondary static group (13.5 Hz vs 9.6 Hz; p=0.002) and higher in the NBIA group than in either the primary (25 Hz vs 13.5 Hz; p=0.006) or the secondary static group (25 Hz vs 9.6 Hz; p=0.00004). The median GPe firing frequency was higher in the NBIA group than in the secondary static group (15.9 Hz vs 7 Hz; p=0.013). The NBIA group also showed a higher proportion of regularly firing GPi cells compared with the other groups (p<0.001). A higher proportion of regular GPi cells was also seen in patients with fixed/tonic dystonia compared with a phasic/dynamic dystonia phenotype (p<0.001). The GPi firing frequency showed a positive correlation with 1-year outcome from DBS measured by improvement in the Burke-Fahn-Marsden Dystonia Rating Scale (BFMDRS-m) score (p=0.030). This association was stronger for the non-progressive patients (p=0.006). Interpretation Pallidal firing rates and patterns differ significantly with dystonia aetiology and phenotype. Identification of specific firing patterns may help determine targets and patient-specific protocols for neuromodulation therapy. Funding National Institute of Health Research, Guys and St. Thomas’ Charity, Dystonia Society UK, Action Medical Research, German National Academic Foundation.

Central Motor Conduction Time and Diffusion Tensor Imaging metrics in children with complex motor disorders

OBJECTIVES To explore potential correlations between Diffusion Tensor Imaging (DTI) metrics and Central Motor Conduction Time (CMCT) in a cohort of children with complex motor disorders. METHODS For a group of 49 children undergoing assessment for potential Deep Brain Stimulation (DBS) surgery, CMCT was derived from the latency of MEPs invoked by transcranial magnetic stimulation of the contralateral motor cortex and from peripheral conduction times. Tract-Based Spatial Statistics (TBSS) was used to compare Diffusion Tensor Imaging (DTI) metrics between children with normal and abnormal CMCT. TBSS was also used to look for correlations between these metrics and CMCT across the group. RESULTS Median age at assessment was 9years (range 3-19years). For 14/49 children a diagnosis of primary dystonia had been made. No correlation could be found between DTI metrics and CMCT, with no difference in metrics found between children with normal and abnormal CMCT. CONCLUSIONS DTI metrics did not differ between children with normal and abnormal CMCT. Tissue properties determining CMCT may not be explained by existing DTI metrics. SIGNIFICANCE DTI and CMCT measurements provide complementary information for the clinical assessment of children with complex motor disorders.

Inconsistent effects of EMG rectification on coherence analysis

The recent article by Farina et al. (2013) makes several important contributions to the controversial debate on whether the raw or rectified electromyogram (EMG) should be used to calculate corticomuscular and intermuscular coherence. They provide clear evidence from both simulated and empirical data that oscillatory inputs to the motoneurone pool can be identified from the raw EMG, and that corticomuscular or intermuscular coherence is detected using the unrectified EMG. Their results therefore concur with Yao et al. (2007) and McClelland et al. (2012). They show further (eqn (11) and Fig. 1) that the effect of rectification varies with differing levels of muscle contraction. This inconsistent effect of rectification, which relates to the non-linear nature of the process, is a concern we have raised previously (McClelland et al. 2012). There are other examples in the literature of rectification enhancing (Ward et al. 2013) or decreasing (Neto & Christou, 2010) coherence, having no effect on coherence (Yao et al. 2007), or creating false coherence where the input has no components (Neto & Christou, 2010). Farina et al. show clearly that the spectral peak relating to the motoneurone pool input is more influenced by the level of muscle activity for the rectified EMG than for the raw EMG (Fig. 1). Furthermore, they show that the coherence peak with respect to the spike train input decreases with increasing numbers of active motor units for the rectified EMG, but, for the raw EMG, remains approximately constant regardless of the level of amplitude cancellation (Fig. 4). These observations are critical, as, for any study analysing corticomuscular coherence during different phases of a motor task, it is vital that the analysis method is consistent across that task. When the rectified EMG is used, it becomes difficult to discern whether a change in coherence magnitude detected between different phases of a task relates to a genuine physiological change or simply reflects the variable effect of rectification with different levels of muscle activation. In contrast, analysis with the raw EMG is able to detect the coherence peak irrespective of the level of muscle activation. From the authors’ theoretical analysis, it further follows that the extent to which rectification will affect the coherence is very difficult to estimate, as the cancellation term c(t) in eqn (11), which expresses the difference between the rectified EMG and the sum of the rectified spike trains, depends in a very complex manner on the input (which is more complex than the sinusoid considered in their analysis). We also raise several concerns regarding some of the analysis and conclusions made in this article. The theoretical analysis neglects the effects of noise. Spinal motoneurones have inputs from multiple sources, resulting in components of the EMG signal which are independent of the considered cortical input. Equation (11) should contain the spectrum of rectified noise and it should be taken into account that the cancellation term will depend on noise in a very complex way. It is primarily this complex interaction, produced by rectification, between noise and the EMG component of interest that makes the effect of rectification on coherence analysis inconsistent. The authors suggest that, at very low contraction levels, rectification may be necessary for the detection of coherence. However, this conclusion is not supported either by their theoretical or experimental analysis. From their eqns (1) and (6), it follows that, assuming that f0i/fm is never a ratio of integers, then at frequency fm the raw EMG has no components other than those due to the cortical input and therefore coherence between the raw EMG and cortical input at that frequency is maximal, i.e. equal to 1. On the other hand, eqn (11) implies that the coherence between the rectified EMG and the cortical input can at most be 1, because of the cancellation term c(t). Hence according to their theoretical analysis, EMG rectification can only decrease coherence (except in the case of a single motor unit train where there is no amplitude cancellation). Their empirical EMG data, recorded at 5–10% maximum voluntary contraction (MVC), show that the coherence peak is clearly detected using the raw EMG. From our own experience, corticomuscular coherence can be detected even when the raw signal power at the relevant frequencies is very low. This reflects the fact that coherence is a measure of the coupling between two signals and not simply a measure of their power (Bruce & Goldman, 1983). Interestingly, a recent article using partial coherence to determine the ability of the surface EMG to account for coherence between pairs of single motor units demonstrated that, although rectification enhanced coherence detection, both the rectified and unrectified EMG predicted the frequency of motor unit synchronization (Ward et al. 2013). Furthermore, there was no difference between the raw and rectified EMG in accounting for single motor unit coherence when the muscle contracted more strongly. Farina et al. raise concerns about the validity of claims that EMG rectification is inappropriate for calculating coherence when those claims are ‘based on signals that do not share the same structure of the EMG’. The theoretical examples we used (McClelland et al. 2012) and the simulations used by Neto and Christou (2010) take a more general signal processing approach than the EMG model used by Farina et al., but this does not make those claims invalid. Furthermore, both our empirical and theoretical analysis provided evidence for the inconsistent and unpredictable effects of EMG rectification on coherence, and these results are supported by Farina et al.’s current analysis. We welcome a rigorous scientific debate on this issue and we congratulate Farina and colleagues on demonstrating the inconsistent effects of EMG rectification. Although the authors suggest that rectification may be necessary in some situations, they correctly point out that it should not be used when comparing coherence across conditions in which amplitude cancellation may vary. This applies to many experimental paradigms investigating motor control. Although EMG rectification is valuable in some forms of analysis, in our opinion there is no good evidence for its application in the calculation of corticomuscular coherence. Corticomuscular coherence can clearly be detected without rectifying the EMG. Even if rectification enhances coherence in some circumstances, the fact that it does so inconsistently is a fundamental problem in scientific study design.

Corticomuscular Coherence With Time Lag With Application to Delay Estimation

Functional coupling between the motor cortex and muscle activity is usually detected and characterized using the spectral method of corticomuscular coherence (CMC). This functional coupling occurs with a time delay, which, if not properly accounted for, may decrease the coherence and make the synchrony difficult to detect. In this paper, we introduce the concept of CMC with time lag (CMCTL), that is the coherence between segments of motor cortex electroencephalogram (EEG) and electromyography (EMG) signals displaced from a central observation point. This concept is motivated by the need to compensate for the unknown delay between coupled cortex and muscle processes. We demonstrate using simulated data that under certain conditions the time lag between EEG and EMG segments at points of local maxima of CMCTL corresponds to the average delay along the involved corticomuscular conduction pathways. Using neurophysiological data, we then show that CMCTL with appropriate time lag enhances the coherence between cortical and muscle signals, and that time lags which correspond to local maxima of CMCTL provide estimates of delays involved in corticomuscular coupling that are consistent with the underlying physiology.

EMG rectification has inconsistent effects on coherence analysis even in single motor unit studies

to the editor: We read with interest the article by [Ward et al. (2013)][1]. Several groups question the use of electromyogram (EMG) rectification in calculating corticomuscular coherence ([McClelland et al. 2012][2]; [Neto and Christou 2010][3]; [Stegeman et al. 2010][4]), and [Farina et al. (2013

Somatosensory Evoked Potentials and Central Motor Conduction Times in Children with Dystonia and their correlation with outcomes from Deep Brain Stimulation of the Globus pallidus internus

Highlights • A high proportion (47%) of children with dystonia have evidence of abnormal sensory pathway function.• Central motor conduction times (CMCTs) and somatosensory evoked potentials (SEPs) show a significant relationship with deep brain stimulation (DBS) outcome, independent of aetiology or cranial MRI.• CMCTs and SEPs can guide patient selection and help counsel families about potential benefit of DBS.

The Neurophysiology of Paediatric Movement Disorders

Purpose of review To demonstrate how neurophysiological tools have advanced our understanding of the pathophysiology of paediatric movement disorders, and of neuroplasticity in the developing brain. Recent findings Delineation of corticospinal tract connectivity using transcranial magnetic stimulation (TMS) is being investigated as a potential biomarker for response to therapy. TMS measures of cortical excitability and neuroplasticity are also being used to investigate the effects of therapy, demonstrating neuroplastic changes that relate to functional improvements. Analyses of evoked potentials and event-related changes in the electroencephalogaphy spectral activity provide growing evidence for the important role of aberrant sensory processing in the pathophysiology of many different movement disorders. Neurophysiological findings demonstrate that children with clinically similar phenotypes may have differing underlying pathophysiology, which in turn may explain differential response to therapy. Neurophysiological parameters can act as biomarkers, providing a means to stratify individuals, and are well suited to provide biofeedback. They therefore have enormous potential to facilitate improvements to therapy. Summary Although currently a small field, the role of neurophysiology in paediatric movement disorders is poised to expand, both fuelled by and contributing to the rapidly growing fields of neuro-rehabilitation and neuromodulation and the move towards a more individualized therapeutic approach.

Delay estimation between EEG and EMG via coherence with time lag

The traditional way to estimate the time delay between the motor cortex and the periphery is based on the estimation of the slope of the phase of the cross spectral density between motor cortex electroencephalogram (EEG) and electromyography (EMG) signals recorded synchronously during a motor control task. There are several issues that could make the delay estimation using this method subject to errors, leading frequently to estimates which are in disagreement with underlying physiology. This study introduces cortico-muscular coherence with time lag (CMCTL) function and proposes a method for estimating the delay based on finding its local maxima. We further address the issue of the interpretation of such time delay in multi-path propagation systems. Delay estimates obtained using the proposed method are more consistent compared with results obtained using the phase method and in a better agreement with physiological facts.