It can be shown in this work that distinct patterns of cortico-muscular and/or intermuscular coherence can be identfied in a variety of movement disorders (cortical myoclonus, limb dystonia, myoclonus of CBD). Additionally, it could be demonstrated that the assessment of the reticulospinal system is feasible by using intermuscular frequency analysis of homologous muscles, which might open up a new line of research of subcortical drives within the motor system.
In high frequency rhythmic myoclonus of cortical origin EEG-EMG and EMG-EMG frequency analysis has potential diagnostic value for distal muscles in providing a simple and sensitive measure of the strength of functional coupling between cortex and muscle. Detailed consideration of EEG-EMG and EMG-EMG phase spectra also provides important information regarding the mechanisms underlying myoclonic bursts in different muscles. In the hand, phase suggests that efferent pathways dominate and jerking seems to be an expression of spontaneous cortical discharges that are intrinsically rhythmic. In more proximal muscles phase relationships may be dominated by either efferent or afferent loops arguing that myoclonus may arise from spontaneous rhythmic cortical discharges or self-sustaining myoclonic activity through afferent-efferent loops.
In patients with limb dystonia due to a variety of etiologies there were relevant differences among etiologies. 10 out of 12 (83%) of symptomatic DYT1 patients had an excessive 4-7 Hz common drive to TA evident as an inflated coherence in this band. This drive also involved GC leading to co-contracting EMG bursts. In contrast, asymptomatic DYT1 carriers, patients with symptomatic dystonia, patients with fixed dystonia and healthy subjects showed no evidence of such a drive in the theta-frequency band nor any other distinguishing electrophysiological feature. Moreover, the pathological 4-7 Hz drive in symptomatic DYT1 patients was much less common in the upper limb, where it was only present in two out of six (33%) of patients with clinical involvement of the arms. It can therefore be concluded that the nature of the abnormal drive to dystonic muscles may vary according to the muscles under consideration and, particularly, with aetiology.
Patients with the myoclonus of CBD can exhibit dramatically inflated EMG-EMG coherence in the absence of any evidence of a pathological cortico-spinal drive as determined by EEG-EMG coherence, raising the possibility of involvement of subcortical motor systems in the myoclonus of CBD. However, given the relatively small size of the sample, more research is needed to define how representative the present findings are in patients with of CBD.
Intermuscular frequency analysis of muscle bursts elicted by the acoustic startle response demonstrated autospectral peaks at around 14 Hz in deltoid and biceps muscles only. Similarly, coherence spectra of the EMG recorded between homologous proximal upper limb muscles demonstrated a peak centred around 12-16 Hz during reflex startles. Coherence in the 10-20 Hz band was significantly greater in the startle reflex than during voluntary sham startles or voluntary tonic contraction for deltoid, but not first dorsal interosseous, muscles. Thus, the coherence at 10 to 20 Hz between EMGs from homologous muscles represents a potential surrogate measure of reticulospinal activity that may be useful in determining the contribution of the reticulospinal system to different types of movement in health and disease.
So far studies of the coherence between cortical activity and EMG or between EMG signals have focussed on long records of essentially stationary physiological activity, such as voluntary tonic contraction or records of persistent tremor. However, these paradigms are relatively limited. Many pathological conditions, such as hyperekplexia and paroxysmal dystonia, lead to involuntary muscle contractions that are brief. Wider adoption of MAR models may permit the determination of the pattern of descending drives in such conditions in the future. In other pathological conditions such as chorea, involuntary movement may be persistent, but vary in an unpredictable fashion.
In these more complex cases, it is not appropriate to apply the standard stationary FFT based spectral estimation techniques. For these types of signal, non-stationary models can capture much more of the true structure of the data. Non-stationary signals are those whose statistical moments, such as the mean and variance change in time through the signal. One [page 91↓]way of approaching non-stationary signals is to consider them as being composed of a number of smaller stationary states in which the statistical properties stay fairly constant. There are a number of standard models that can be employed to probabilistically determine these smaller stationary regimes (or states) and their respective spectral properties. Such an approach is particularly suitable for objectively segmenting signals into regimes corresponding to different states of muscle activation and rest (Cassidy and Brown, 2002). Periods of stationary activity detected on probabilistic grounds can then be averaged for better spectral estimates. This approach may therefore prove useful in the determination of the pattern of descending drive in conditions such as chorea. Alternatively, one can employ a model whose properties change dynamically though the data record. Such an approach would be more suitable where signals change gradually so that discrete state change times are hard to discern, as in event-related (de) synchronisation paradigms.
Advances need not be solely analytical. More work is necessary on the pharmacological underpinning of cortico-muscular and intermuscular coherence through the systematic investigation of drug effects and ligand-gated channelopathies, and normal ranges for EEG-EMG and EMG-EMG coherences at different frequencies clearly need to be established. Thus, at present there is still a considerable way to go before frequency analysis can provide an accessible and useful tool in the assessment of disorders of the motor system in a routine clinical practice.
|© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.|
|DiML DTD Version 3.0||Zertifizierter Dokumentenserver|
der Humboldt-Universität zu Berlin