Exp Brain Res (2002) 145:207–214 DOI 10.1007/s00221-002-1113-x
R E S E A R C H A RT I C L E
R.J. Fisher · M.P. Galea · P. Brown · R.N. Lemon
Digital nerve anaesthesia decreases EMG-EMG coherence in a human precision grip task
Received: 26 November 2001 / Accepted: 13 March 2002 / Published online: 16 May 2002 © Springer-Verlag 2002
Abstract There is increasing evidence that the primary motor cortex is involved in the generation of electromyographic (EMG) oscillations at frequencies in the range of 15–30 Hz that are observed during performance of a precision grip task. Since the level of the corticomuscular coherence varies according to the nature of the object that is gripped, it seemed possible that somatosensory inputs from the hand might affect this coherence. The aim of this study was to investigate whether interrupting cutaneous inputs from the digits would affect the coherence between hand muscles during precision grip of a compliant object. Subjects performed a precision grip hold-ramphold task before, during and after digital nerve anaesthesia of the index finger and thumb. There were marked deficits in the performance of the task, particularly during the initial formation of the grip and first hold period. Local digital nerve anaesthesia reduced but did not abolish 14–31 Hz coherence between EMG activity recorded from different hand and forearm muscles. Coherence was measured during the second hold phase of the task. Digital nerve anaesthesia did not affect the predominant frequencies in the EMG power spectra compiled from the same phase of the task. We conclude that during a precision grip task, cutaneous input enhances oscillatory synchrony between pairs of hand muscles. Keywords Oscillations · Coherence · Anaesthesia · EMG · Synchrony
Introduction It is now widely accepted that electromyographic (EMG) oscillations in the 15–30 Hz range recorded during a precision grip task, are at least in part, driven from central R.J. Fisher · M.P. Galea · P. Brown · R.N. Lemon (✉) Sobell Department of Motor Neuroscience and Movement Disorders, Institute of Neurology, Queen Square, London WC1N 3BG, UK e-mail:
[email protected] Tel.: +44-20-78373611 Ext. 4184, Fax: +44-20-78133107
structures, including the primary motor cortex. This conclusion was based on recordings of motor cortex oscillations in the 15–30 Hz range in monkeys (Murthy and Fetz 1992; Sanes and Donoghue 1993; Baker et al. 1997) and measurements in humans of coherence between cortical signals and EMG activity (Conway et al. 1995; Salenius et al. 1997a; Mima et al. 2000). Coherence between magnetoencephalographic (MEG) and EMG signals and between EMG signals from different hand muscles is particularly pronounced during the steady hold phase of the precision grip task, but is abolished during movement (Baker et al. 1997; Kilner et al. 1999, 2000). Recently it was also shown that both MEG-EMG and EMG-EMG coherence in the 15–30 Hz range is modulated by the parameters of the precision grip task (Kilner et al. 2000), with increased coherence observed when subjects gripped more compliant objects. It was concluded that levels of coherence in the hold phases of this task might reflect important changes in sensorimotor state encompassing alterations in both grip force and digit position. Cutaneous inputs from the digits are an important source of sensory events controlling precision grip, and provide essential information about object compliance (for review, see Johansson 1996). Thus it was of interest to investigate whether interrupting these inputs would affect the coherence between hand muscles during precision grip of a compliant object. Although such coherence has been considered to originate mainly from centrally generated rhythms (Salenius et al. 1997a; Hari and Salenius 1999; Mima et al. 2000) it can be influenced by somatosensory input. For example, it has been shown that median nerve stimulation provoked an immediate reduction, then an increase or rebound in the amplitude of sensorimotor 10 and 20 Hz mu rhythms (Salmelin and Hari 1994; Salenius et al. 1997b). We focused on changes in the ongoing level of EMG-EMG coherence that occur after a movement into a new steady hold period, since this transition is accompanied by a particularly marked increase in coherence that might involve sensory feedback from the hand.
208
Fig. 1A–D Details of the experimental set-up. Subjects performed a precision grip task by squeezing two levers (A) and following target boxes displayed on a computer screen (B). Lever position indicators were provided as feedback (B), and their position in time was also recorded online (single trial shown; C). Subjects performed a hold-ramp-hold task and electromyographic (EMG) recordings were taken from first dorsal interosseous (1DI), abductor pollicis brevis (AbPB), flexor digitorum superficialis (FDS) and extensor digitorum communis (EDC) muscles. Data from a single trial for the hold 2 period is shown in D. Coherence analysis was performed on hold 2 data only (marked with box; C)
Materials and methods
The aim of this study was therefore to investigate the effects of digital nerve anaesthesia on behaviour and EMG oscillatory activity and coherence during a precision grip hold-ramp-hold task.
Behavioural task
Subjects Ten healthy human subjects (six male and four female), all righthanded and aged between 20 and 55 years, were recruited for these experiments. All gave informed consent. The project was approved by the Joint Medical Ethics Committee of the National Hospital for Neurology and Neurosurgery and the Institute of Neurology and was carried out in accordance with the Declaration of Helsinki.
Subjects performed a precision grip task by squeezing two aluminium levers between thumb and index finger of the right hand (Fig. 1A). The levers (20 mm wide by 40 mm long) were housed
209 in a manipulandum placed on a table in front of the subject and were mounted on the axes of two DC motors that were controlled by a robotic device (Phantom Haptic Interface; Sensable Devices, Woburn, Mass., USA). This device simulated a compliant or springy lever, having a spring constant of 0.05 N/mm at the tip of the lever. The task involved an initial hold period (hold 1) for 3 s at a steady force level of 1.3 N, a slow increase in grip force by 0.3 N over 2 s (ramp) to another 3-s hold period (hold 2) (Fig. 1C). The first hold period required a displacement of each lever by 6 mm (ca 8.5° at the DC motor shaft), and a further 6 mm displacement was needed to reach hold 2. Visual feedback of the lever positions was provided by cursors on a computer monitor (Fig. 1B), and subjects were instructed to keep these cursors within two target boxes, also displayed on the monitor. Each new trial began when the target boxes appeared on the monitor screen in the hold 1 target position. The intertrial interval was 2 s. A screen blocked the subject’s view of their hand. Index finger and thumb lever positions were recorded using optical encoders (resolution of ~33 counts per mm movement of the lever tip). EMG and lever position recordings Bipolar surface EMG recordings (Fig. 1D) were taken for the first dorsal interosseous (1DI), abductor pollicis brevis (AbPB), flexor digitorum superficialis (FDS) and extensor digitorum communis (EDC) muscles. EMGs were amplified (×1,000), high-pass filtered (30 Hz) and sampled at 5,000 Hz by a PC-compatible computer attached to a 1401+ interface (Spike 2; CED, Cambridge, UK). Lever position signals were sampled at 1 kHz. Digital anaesthesia Local ring block anaesthesia of the thumb and index finger was performed by the injection of 1 ml lignocaine hydrochloride (1% w/v; Antigen Pharmaceuticals) into each side of the base of the proximal phalanx of each digit (2 ml per digit). Prior to the induction of digital anaesthesia the threshold of sensation to electrical stimuli of the tips of the index finger and thumb was determined. For this test small rectangular surface electrodes (40×26 mm) were firmly attached to the lateral borders of the two digits along the path of the digital nerves. Electrical stimuli (Digitimer stimulator DS7) were delivered at a randomised rate of approximately 0.5 Hz (0.2–0.7 mA, 100 µs pulse width) for 1–2 min, and subjects were asked to report the occurrence of all detectable stimuli. The level of local anaesthesia induced by lignocaine was tested clinically using light touch stimuli using cotton wool, as well as by the threshold to electrical stimulation applied to the digit tips. On most occasions the doubling of threshold to electrical stimulation marked the point at which subjects were clearly unable to detect light touch.
ments during the hold 2 phase of the task; such trials were excluded. In order to have equal trial numbers for each condition, the first 60 trials from each experimental block were then used for analysis. Behavioural analysis Index finger and thumb lever velocities were calculated by differentiating the recorded lever position signals. The velocity profile of each lever for each trial was then smoothed using a 50-point Hanning window to remove small high-frequency signals due to the encoder digitisation process. It was then rectified. The average rectified velocity during a 2 s period in each of the three phases of the task (hold 1, ramp and hold 2) for each trial was calculated; the last 2 s of each hold phase was used. These mean velocity values were then averaged over all 60 trials for each phase of the task. Statistics Significant differences in velocity of index finger or thumb levers between pre-anaesthesia and anaesthesia trials, and between preanaesthesia and washout trials, were tested with a one tailed, paired t-test run across subjects. Our a priori hypothesis was that local anaesthesia would cause a reduction in coherence levels, based on the fact that lowering the level of compliance of an object has been shown to cause a significant reduction in coherence (Kilner et al. 2000). Frequency-domain analysis EMG signals were rectified and down-sampled to an effective sampling rate of 263 Hz. Power spectra for 1DI and AbPB muscles, and coherence between all muscle pairs, were calculated as described by Farmer et al. (1993) and Baker et al. (1997). Fast Fourier Transforms of the EMG data were performed using 4×128 point non-overlapping rectangular windows (permitting a 2-Hz resolution). Frequency-domain analysis was performed on the last 2 s of EMG data from the hold 2 phase of the task only. For the 1DI/AbPB muscle pair, coherence values between 14 and 31 Hz were summed for each subject and then multiplied by the bin width (2 Hz) to calculate the area of coherent activity. In order to calculate an overall coherence value from all muscle pairs of each subject, coherence values for muscle pairs had to be normalised. This was achieved by using a transformation as described previously (Kilner et al. 1999). The calculation used was: (1) where C is coherence between muscle pairs 1 and 2 and N is the number of disjoint sections used (4 windows×60 trials=240). For a given subject, transformed muscle pair coherence values at each frequency were then combined as follows:
Experimental protocol (2)
Subjects performed 65 trials prior to (“pre-anaesthesia” trials) and at least 70 trials after the induction of digital anaesthesia (“anaesthesia” trials). Anaesthesia trials were started as soon as there was a complete loss of light touch sensation from the digit tips and/or a doubling of the threshold to electrical stimulation. Subjects performed a further 65 trials approximately 2 h later once the anaesthesia had worn off (“washout” trials), as determined by the return of light touch sensation and a return of the threshold to electrical stimulation to baseline levels. The relatively short duration of digital anaesthesia and length of the experimental protocol precluded detailed somatosensory testing.
where M is the number of muscles used and P is the number of muscle pairs combined. Combined values for each subject were then summed within the 14–31 Hz frequency range. This value was then multiplied by the frequency bin width (2 Hz) to obtain the “coherence area” for all muscle pairs in that subject. For technical reasons, the muscle pair FDS/EDC was excluded from the analysis of all muscle pair coherence.
Data analysis
Coherence larger than S may be considered significantly different from zero with P
