Physiological tremor is an involuntary oscillatory movement in the distal parts of the body. However, physiological tremors are no longer considered as biological noises. The peripheral manifestations are ascribed to the cortical rhythm of the motor circuits and spinal reflexes (8,23) in regulation for in-phase rhythms of the motor unit (23). Recent studies have shown that examining the intersegmental dynamics of physiological tremor could provide insight into a fine level of manipulative control, such as visuomotor tracking (14,39) and postural maintenance (26-28). For instance, finger movement fluctuations during unsupported postural holding can be minimized with active compensatory synergy of the wrist joint, as evidenced by an out-of-phase link between segment tremors across the wrist joint (27,28). Because of the enhanced interdigit tremor coupling at 8-12 Hz (15), repetitive contraction of a single finger causes the spread of central fatigue, undermining finger-fractionated movement. In contrast, load superimposition on a single finger leads to a global decrease in interdigit tremor coupling, representing increased digit independence during manual tracking (11).
Motor overflow refers to covert muscle activity that may accompany the production of voluntary movement (1,10). Motor overflow can even produce cross excitation of the muscles of one side of the body, when mirror reversals on the opposite side are performed. After a period of unilateral resistance training, strength transfer in the untrained limb due to motor overflow has been validated, similar to skill transfer between limbs (7). Neural mechanisms of both central and peripheral origins are hypothesized (5,21), including an altered transcallosal process (10,21) and modulation of an unidentified oligosynaptic or polysynaptic spinal pathway across the midline (5). Motor overflow becomes manifest when the task complexity or exertion level of a contralateral paradigm increases (1). However, cross transfer of motor excitation does not always provide functional benefits because motor overflow spreads globally, causing coupled activation of homologous and nonhomologous muscles (4,10,13). Excessive cross excitation may therefore place neural constraints on the manipulative dexterity of the opposite limb (36), especially in the case of pianists, who must perform bimanually to achieve expert proficiency (33). However, most studies concerning the motor characteristics of pianists have been confined mostly to either keystroke dynamics for a strict serial execution (22,25) or structural adaptations in the cortical or neuromuscular systems (30,31). The effort-related lateral spread of excitation on behavior variables for "superathletes" has not undergone experimental evaluation up to now.
By examining the interplay of multidigit tremors, this study contrasts the effect of contralateral motor overflow on finger independence in pianists and nonmusician controls. The design of the experiment was such that subtle changes in the group-dependent spread of motor overflow could be characterized by different tremor features and interdigit tremor coupling. It was hypothesized that pianists would be able to develop suppressive control over movement fluctuations caused by motor overflow. Adaptive tremor organization and potential neural correlates against cross excitation due to intensive music practice are discussed.
Twenty-four healthy subjects participated in this study. They were divided into two groups. The pianist group was composed of 12 college-aged piano students (18-23 yr; 10 males and 2 females) with regular piano practice for 11-16 yr. The control group consisted of 12 age-matched nonmusicians (19-28 yr; 9 males and 3 females) recruited from a local community and a university. All of the subjects were self-reported as being right-handed, and none of them had symptoms or signs of neuromuscular diseases. The research project was approved by an authorized institutional human research review board, and all subjects signed informed consents before the experiment.
All subjects conducted two contralateral resistance protocols in the nongrip and gripped conditions for three times each in a randomized order with the left hand. The subject sat comfortably on a chair with his/her dominant (right) wrist and forearm fixed within a thermoplastic splint on the table (Fig. 1A). The left (nondominant) arm hung naturally by the trunk. In the grip condition, the subjects were asked to curl the left fingers into the palm and retract the left thumb with maximal effort for 30 s. The maneuver of the nondominant hand was expected to introduce a greater contralateral motor overflow than that of the dominant hand (4). In the nongrip condition, the subjects just lightly clenched their left fist. During contralateral resistance protocols, the subjects spread all their fingers of the right hand in parallel to the ground, pointing the extended fingers ahead without any support. Physiological tremors in the vertical direction of the right-hand digits were measured with four accelerometers (ADXL203; Analog Devices, Inc., Austin, TX; sensitivity = 1000 mV·g−1, measurement range = ±1.7 g, bandwidth = 2.5 kHz). The accelerometers were mounted on the dorsal aspect of the distal phalanges of the right index, middle, ring, and little fingers. Finger oscillatory activities were recorded and then amplified with a gain of 10. Muscle activity of the right extensor digitorum (ED) and the right flexor digitorum superficialis (FDS) was recorded by a pair of bipolar surface electrode units (1.1 cm in diameter, gain = 365, common mode rejection ratio (CMRR) = 102 dB; Imoed, Inc., Salt Lake City, UT). The electrode on the right ED was placed at three-quarters of the distance from wrist to elbow. The muscle activity of the right FDS was recorded by placing the electrode at an oblique angle approximately 4 cm above the wrist on the palpable muscle mass. Sample data of digit tremors and muscle activities of the FDS/ED muscles of the unexercised right hands of a nonmusician and a pianist are shown in Figure 1B. All the signals were sampled at 1 kHz using a custom program on a LabVIEW platform (LabVIEW v.8.5; National Instruments, Austin, TX).
Data processing and reduction.
Accelerometer data and EMG of the ED and FDS muscles were conditioned with band-pass filters (pass band for accelerometers = 1-40 Hz; pass band for EMG = 1-400 Hz), so that a linear trend of a biased current or crosstalk of cardioballistics would be precluded before feature extraction. The amplitude of the EMG of the ED and FDS muscles of the right arm was represented in terms of root mean square (RMS). Principal component analysis with a covariance method was performed within subjects to condense high-dimensional tremor data of an experimental trial for the nongrip and grip conditions (15). The tremor data of four digits were transformed into two principal components (PC1 and PC2) in the form of a time series. These two components could account for >80% of the total tremor variance properties of an experiment trial in the grip or nongrip condition. The communality (hPC12, hPC22) of tremor PC was the square values of correlation coefficients between the PC1/PC2 and individual digit tremors. In fact, hPC12 and hPC22 determined the amount of digit tremors that contributed to the principal components. The amplitudes of PC1 and PC2 of each trial were calculated by applying an RMS to the signals, and the three PC RMS values were averaged to represent PC amplitudes of a subject in the nongrip and grip conditions. The overflow effect on digit tremors was defined as the difference in PC amplitude between the grip condition and nongrip condition divided by that of the nongrip condition, or (RMSgrip − RMSnongrip)/RMSnongrip. To visualize spectral organizations of PC, we used the Welch method to estimate the frequency components PC of each trial. Three trials of PC spectral profile were pooled and averaged for each subject in the nongrip and grip conditions. A Hanning window was applied to each artifact-free 2.56-s epoch of tremor PC with an overlap of 0.64 s, and the spectral density was calculated using a fast Fourier transform. A multiscale entropy area (MSE area) based on sample entropy was used to achieve the quantification of complexity variations in PC of a trial (see Appendix A, Supplemental Digital Content 1, https://links.lww.com/MSS/A86, which demonstrates equations for MSE area calculation) (6). Mean MSE areas of PC1 and PC2 in the nongrip and grip conditions were obtained by averaging MSE values of the three trials within subjects. Mean MSE area has been shown to be a robust biomarker to measure the complexity of biological time series, with a higher MSE area indicating a more complex and noisy structure of tremor PC (6).
Tremor coupling of among digits in the unexercised hand was characterized with mutual information (MI) (Appendix B, Supplemental Digital Content 2, https://links.lww.com/MSS/A87, which demonstrates equations for mutual information calculation). The MI of a given pair of digit tremors can be regarded as a nonlinear equivalent of the correlation function between accelerometer data measured from the fingers. Three MI of a tremor pair in the nongrip and grip conditions were averaged to represent the level of tremor coupling of the digit pair for a subject. Mutual information is nonnegative. If two tremor time series (X and Y) are completely independent, MI equals 0. A high MI represents a strong mutual dependence of two tremor signals.
A repeated-measure two-way ANOVA (resistance protocol (grip, nongrip) × group (pianist, control)) and post hoc analyses were used to examine the effects of contralateral overflow and group on 1) the RMS of EMG of the FDS and ED muscles and 2) the temporal/spectral features of tremor PC, including RMS, and the MSE areas of PC1 and PC2. The group difference of standardized changes in PC amplitudes due to additional resistance protocol in the opposite limb, or (RMSgrip − RMSnongrip)/RMSnongrip, was further examined with independent t-test. For both the pianist and control groups, a Hotelling's T2 statistic was used to contrast the MI of six tremor pairs in the grip and nongrip conditions. The levels of significance for the determination of difference and post hoc analyses were 0.05. Signal processing and statistical analyses were completed using Matlab v.7.0 (MathWorks, Inc., Natick, MA) and the statistical package for Social Sciences (SPSS) for Windows v. 15.0 (SPSS, Inc., Chicago, IL). Data reported in the texts, figures, and tables are all presented as mean ± SE.
The means and SE of EMG RMS of the ED for the pianist group in the grip and nongrip conditions were 11.50 ± 0.05 and 11.70 ± 0.06 μV, respectively. For the control group, EMG RMS of the ED was 17.90 ± 0.24 and 14.28 ± 0.16 μV in the grip and nongrip conditions, respectively. The ANOVA result suggested that the EMG RMS of the ED did not differ between groups (F1,11 = 4.38, P = 0.071), resistance protocols (F1,33 = 1.14, P = 0.293), or the interaction of both (F1,33 = 1.35, P = 0.254). For the FDS, the RMS value of the muscle was subject only to the protocol effect (F1,33 = 4.68, P = 0.031) rather than the group effect (F1,11 = 3.68, P = 0.08) and the interaction effect (F1,33 = 0.88, P = 0.355). Post hoc analysis revealed a greater RMS in the grip condition (6.35 ± 0.03 μV) than in the nongrip condition for the control group (5.41 ± 0.02 μV, P = 0.015). However, the RMS of the FDS for the pianist group in the grip condition (5.73 ± 0.02 μV) did not differ from that of the nongrip condition (5.23 ± 0.01 μV, P > 0.05).
Temporal and spectral features of the PC.
Table 1 shows the overall variance properties explained by the two major principal components (PC1 and PC2) in the grip and nongrip conditions. For both of the groups, PC1 was always the predominant PC, capturing roughly 60% of the variance property of digit tremors. PC1 and PC2 accounted in total for >80% of the variance of tremor data. This fact warranted that we focus on structural changes in PC due to the effects of contralateral motor overflow. Table 2 shows group means of communality (hPC12 and hPC22) that estimate the relative contributions of each digit tremor to PC. For the nonmusician and pianist groups, PC1 generally had higher communality, or hPC12, with digit tremor of the index, middle, and ring fingers (0.440-0.727). In contrast, PC2 exhibited a relatively high communality, or hPC22, with digit tremor of the index finger (0.330-0.382) but with the lowest communality for the middle finger (0.037-0.138) (Table 2).
Figure 2 (A and B) displays group means of the PC amplitude in the grip and nongrip conditions. The ANOVA results suggested a significant group × protocol interaction effect on PC1 RMS (F1,33 = 4.70, P = 0.031). Post hoc analysis indicated that the PC1 amplitude of the pianist group was smaller in the grip condition than in the nongrip condition (P = 0.008). Contrary to the pianist group, the control group demonstrated a greater PC1 amplitude in the grip condition than in the nongrip condition (P = 0.005). In addition, the two groups exhibited opposite trends in PC1 amplitude in the nongrip and grip conditions. The pianist group exhibited a larger PC1 amplitude than the control group in the nongrip condition, whereas the control group had greater PC1 amplitude than the pianist group in the grip condition (P < 0.005). The RMS value of PC2 was not significantly subject to protocol or group effects (group: F1,11 = 0.07, P = 0.798; protocol: F1,33 = 0.25, P = 0.619), or the interaction of both (F1,33 = 2.89, P = 0.098). Figure 2C contrasts the standardized changes in the amplitudes of PC1 and PC2 in the pianist and control groups. The standardized changes in PC amplitude differed for both groups (PC1: t = 5.648, P < 0.001; PC2: t = 3.695, P < 0.005). Contralateral maximal resistance caused positive amplitude changes in PC1 and PC2 for the control group, in contrast to the negative amplitude modulations for the pianist group. Figure 3 contrasts the pooled PC spectral profiles of the grip and nongrip conditions for the pianist and control groups. There was a prominent 8- to 12-Hz spectral peak in PC1 in both the grip and nongrip conditions. ANOVA statistics indicated that the 8- to 12-Hz spectral peak in PC1 was differentially modulated with the resistance protocol for both groups. To examine the significant group × protocol interaction (F1,33 = 4.21, P = 0.048), we conducted further post hoc analysis, and the results suggested an effort-related enhancement of the 8- to 12-Hz spectral peak in PC1 for the control group (P < 0.05). However, for the pianist group, the spectral peak was relatively suppressed in the grip condition compared with that in the nongrip condition (P < 0.05). The spectral distribution of PC2, which consisted of additional higher 15- to 30-Hz tremor components, showed no significant group or protocol differences (group: F1,11 = 0.33, P = 0.576; protocol: F1,33 = 0.18, P = 0.671; group × protocol: F1,33 = 2.17, P = 0.15).
Complexity analysis of PC.
Figure 4 (A and B) displays different effort-induced modulations of the MSE area of the PC1 and PC2 in the control and pianist groups. The ANOVA comparison showed a significant group × protocol interaction effect on the PC1 MSE area (group: F1,11 = 0.09, P = 0.776; protocol: F1,33 = 0.04, P = 0.836; group × protocol: F1,33 = 4.20, P = 0.048). Post hoc analysis suggested that the control group showed a larger PC1 MSE area in the grip condition than in the nongrip condition (P < 0.001), but the opposite trend was noted for the pianist group, with a comparatively lower PC1 MSE area in the grip condition (P = 0.002) (Fig. 4A). However, ANOVA results did not show any significant group or protocol effects on the PC2 MSE area (group effect: F1,11 = 0.12, P = 0.739; task effect: F1,33 = 0.09, P = 0.762; interaction: F1,33 = 2.58, P = 0.118) (Fig. 4B).
Tremor MI among digit pairs.
Figure 5 (A and B) displays the topological organizations of the tremor MI for all digit pairs in the nongrip and grip conditions. In the nongrip condition, the MI for both of the groups was, in general, organized with a decreasing trend as the interdigit distance increased. For the control group, the spatial pattern of MI in the grip condition was similar to the nongrip condition, except that the MI values were intensified compared with those in the nongrip condition (Λ = 0.42, P = 0.001) (Fig. 5A). For the pianist group, MI values of digit tremor were, conversely, lower in the grip condition than those in the nongrip condition (Λ = 0.313, P = 0.018), especially for MI23, MI24, and MI45 (P < 0.05) (Fig. 5B).
This study was the first to characterize pianist-control differences in the cross transfer of motor overflow with physiological tremor. The most important finding from this experiment was that, when contralateral strenuous gripping was conducted, pianists exhibited smaller amplitudes and less complexity of digit tremors, and less interdigit tremor coupling, than musically untrained controls. These findings suggest that pianists can suppress the lateral spread of motor overflow, which may place a neural constraint on finger fractionated movement, thereby allowing complex accomplishments in bimanual dexterity.
Differential cross effect on digit tremors between groups.
The group difference in the cross effect on digit tremors was ascribed primarily to the modulation of PC1, which explained the majority of variances of all digit tremors. Contralateral gripping potentiated PC1 of the nonmusician controls, but PC1 of the pianists was, conversely, suppressed (Fig. 2A). The resulting potentiation of PC1 for the nonmusicians was attributable to known contralateral motor overflow, the amount of which is proportional to the increase in muscular contraction against high resistance (10,12). However, such an effort-dependent PC1 enhancement for the nonmusicians could not be ascribed to strength differences in gripping between the two groups. The pianist-control difference in maximal pinch and grasp forces is insignificant, and piano training alters the acuity of moving fingers rather than the muscle strength (3,31). On the other hand, the pianist group was noted to have an intriguingly greater PC1 than the control group in the nongrip condition. Although a person who is good at very dexterous work is not expected to demonstrate a high level of tremor in a postural task, our observation was congruent with the finding of Walsh (40), who reported a potentiated physiological tremor of a postgraduate student majoring in piano. A possible explanation for the atypical observation could be that musicians might unconsciously develop a voluntary tremor (or vibrato movement) of the arm and hand in playing positions, very similar to involuntary physiological tremor (34). But the rate and amplitude of vibrato movement can be modulated by musicians at will or by feedback for behavioral contexts, such as instrument playing and voluntary exertion. In this study, we also noted a substantial waning of digit tremor in pianists during contralateral gripping (Fig. 2A).
The between-group difference in the cross effect was most likely to be a consequence of training-induced neural adaptation that altered tremor genesis at the supraspinal level. Studies have shown that the coding of the 8- to 12-Hz rhythm in the motor cortex can synchronize different motor networks (23) and modulate muscle activity in a similar form of tremulous oscillation through descending corticospinal pathways (23). Hence, PC1, with a prominent 8- to 12-Hz spectral peak in this study, symbolizes a common central drive to all uninstructed fingers that is amendable to contralateral exertion (Fig. 3). For the nonmusician controls in the grip condition, motor overflow increased sharing of PC1 (Fig. 2A) and homologous activation of the unexercised FDS resulting from transcallosal facilitation (10,32). The cortical activation in the contralateral hemisphere in turn reinforces coinciding excitation of the same cortical region in the ipsilateral hemisphere (1,38). However, piano players were immune to the cross excitation that led to PC1 augmentation in the nonmusicians. When the contralateral hand remained active during strenuous gripping, pianists could simultaneously suppress cross-excited common input to different subdivisions of the unexercised digit muscles. A recent study using transcranial magnetic stimulation provided supporting evidences for this argument that the musicians could reduce transcallosal facilitation and intracortical facilitations within the ipsilateral motor cortex (29). Also, during low-level isometric contraction, musicians exhibit a greater discharge variability of motor units (or less short-term synchrony) than those in untrained subjects and weightlifters (35). In this regard, piano players could flexibly perform a wide variety of output combinations on the keyboard (16) without being constrained by motor overflow. When inhibitory modulation against cross excitation is impaired, piano players suffer from mirror dystonia (37). When they play bimanually, this affliction causes performance deficits for inappropriately sequenced movements and undesirable muscle coactivation.
Recent imaging studies have shown that nonmusicians exhibit stronger bilateral cortical and subcortical responses than professional pianists during unimanual finger movements, including the cerebellum (20), somatosensory area, primary motor cortex, and supplementary motor area (9,17). Hence, piano players are considered to adopt more economical central processing than nonmusicians do during keyboard playing by inhibiting superfluous cortical activation and inadvertent contralateral activity (20). As those areas that exhibit enhanced excitation for nonmusicians constitute neural circuit of tremor genesis (23), we may well argue that surplus recruitment of motor circuits in the nonmusicians is likely to contribute to a cross enhancement of tremor complexity (Fig. 4A). Musically untrained controls might invest more unnecessary neuronal activities from different tremor association areas, thereby increasing the complexity of PC1 during contralateral gripping.
Digit independency and interdigit tremor coupling due to motor overflow.
Characterized by the amount of MI, coupling between digit tremors is valuable in the exploration of nonlinear constraints on finger movements. In effect, completely fractionated movement of the fingers for nonmusicians and pianists is never possible. Finger movement can be physically hampered by 1) simultaneous actions of the fingers distributed from tension exerted at one point of proximal aponeurosis or 2) short-term synchrony between motor units acting on different fingers (18,19), known as mechanical and neural constraints, respectively. Keen and Fuglevand (18) reported that the degree of synchronization within compartments of the ED muscle (middle-ring > ring-little > index-middle, middle-little) was in good agreement with the patterned organization of MI between digit tremors in the nongrip condition (Fig. 5, A and B). When maximal effort was made in the opposite hand, the strength of MI was substantially elevated for the nonmusician group, implying a relative degree of motion enslavement among fingers. Enhanced interdigit coupling would decrease the control capacity and independent use of fingers for nonmusicians. A similar finding of an enslaving action of fingers has been reported for the spread of central fatigue after repetitive contraction of a single finger (15). In contrast, because the cross effect on interdigit tremor coupling was inhibitory, pianists are thought to have a greater biomechanical degree of freedom that allows them to accomplish vastly complex finger movements when they play keyboards bimanually.
In terms of multidigit tremor dynamics, piano players were noted to demonstrate an important response inhibition against the cross excitation due to motor overflow that was typically found in nonmusician controls. The differential cross modulation is a compensatory event of intensive keyboard practice, relevant to reduction in the common oscillatory drive of central origins to the unexercised fingers. Our finding highlights the fact that suppression of the lateral spread of unintended motor overflow serves as a crucial basis for piano players to have greater segregated control over finger movements when the opposite hand is working.
This research was supported by grants from the National Science Council, ROC (NSC-94-2314-B-006-026).
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Abercrombie MLJ, Lindon RL, Tyson MC. Associated movements in normal and physically handicapped children. Dev Med Child Neurol
3. Aoki T, Furuya S, Kinoshita H. Finger-tapping ability in male and female pianists and nonmusician controls. Motor Control
4. Armatas CA, Summers JJ, Bradshaw JL. Handedness and performance variability as factors influencing mirror movement occurrence. J Clin Exp Neuropsychol
5. Carroll TJ, Herbert RD, Munn J, Lee M, Gandevia SC. Contralateral effects of unilateral strength training: evidence and possible mechanisms. J Appl Physiol
6. Costa M, Goldberger AL, Peng CK. Multiscale entropy analysis of complex physiologic time series. Phys Rev Lett
7. Farthing JP, Chilibeck PD, Binsted G. Cross-education of arm muscular strength is unidirectional in right-handed individuals. Med Sci Sports Exerc
8. Hagbarth KE, Young RR. Participation of the stretch reflex in human physiological tremor. Brain
9. Haslinger B, Erhard P, Altenmüller E, et al. Reduced recruitment of motor association areas during bimanual coordination in concert pianists. Hum Brain Mapp
10. Hoy KE, Fitzgerald PB, Bradshaw JL, Armatas CA, Georgiou-Karistianis N. Investigating the cortical origins of motor overflow. Brain Res Brain Res Rev
11. Hwang IS. Role of load-induced reorganization of multi-digit physiological tremors for a tracking maneuver. Eur J Appl Physiol
12. Hwang IS, Abraham LD. Quantitative EMG analysis to investigate synergistic coactivation of ankle and knee muscles during isokinetic ankle movement. Part 1: Time amplitude analysis. J Electromyogr Kinesiol
13. Hwang IS, Tung LC, Yang JF, Chen YC, Yeh CY, Wang CH. Electromyographic analyses of global synkinesis in the paretic upper limb after stroke. Phys Ther
14. Hwang IS, Wu PS. The reorganization of tremulous movements in the upper limb due to finger tracking maneuvers. Eur J Appl Physiol
15. Hwang IS, Yang ZR, Huang CT, Guo MC. Reorganization of multidigit physiological tremors after repetitive contractions of a single finger. J Appl Physiol
16. Inui N, Ichihara T. Comparison of the relation between timing and force control during finger-tapping sequences by pianists and non-pianists. Motor Control
17. Jäncke L, Shah NJ, Peters M. Cortical activations in primary and secondary motor areas for complex bimanual movements in professional pianists. Brain Res Cogn Brain Res
18. Keen DA, Fuglevand AJ. Common input to motor neurons innervating the same and different compartments of the human extensor digitorum muscle. J Neurophysiol
19. Kilbreath SL, Gandevia SC. Limited independent flexion of the thumb and fingers in human subjects. J Physiol
20. Koeneke S, Lutz K, Wüstenberg T, Jäncke L. Long-term training affects cerebellar processing in skilled keyboard players. Neuroreport
21. Lee M, Carroll TJ. Cross education: possible mechanisms for the contralateral effects of unilateral resistance training. Sports Med
22. Loehr JD, Palmer C. Sequential and biomechanical factors constrain timing and motion in tapping. J Mot Behav
23. McAuley JH, Marsden CD. Physiological and pathological tremors and rhythmic central motor control. Brain
25. Minetti AE, Ardigò LP, McKee T. Keystroke dynamics and timing: accuracy, precision and difference between hands in pianist's performance. J Biomech
26. Morrison S, Newell KM. Inter- and intra-limb coordination in arm tremor. Exp Brain Res
27. Morrison S, Newell KM. Postural and resting tremor in the upper limb. Clin Neurophysiol
28. Morrison S, Newell KM. Limb stiffness and postural tremor in the arm. Motor Control
29. Nordstrom MA, Butler SL. Reduced intracortical inhibition and facilitation of corticospinal neurons in musicians. Exp Brain Res
30. Oztürk AH, Tasçioglu B, Aktekin M, Kurtoglu Z, Erden I. Morphometric comparison of the human corpus callosum in professional musicians and non-musicians by using in vivo
magnetic resonance imaging. J Neuroradiol
31. Penn IW, Chuang TY, Chan RC, Hsu TC. EMG power spectrum analysis of first dorsal interosseous muscle in pianists. Med Sci Sports Exerc
32. Post M, Bayrak S, Kernell D, Zijdewind I. Contralateral muscle activity and fatigue in the human first dorsal interosseous muscle. J Appl Physiol
33. Rosenkranz K, Butler K, Williamon A, Rothwell JC. Regaining motor control in musician's dystonia by restoring sensorimotor organization. J Neurosci
34. Schlapp M. Observations on a voluntary tremor-violinist's virbato. Q J Exp Physiol Cogn Med Sci
35. Semmler JG, Sale MV, Meyer FG, Nordstrom MA. Motor-unit coherence and its relation with synchrony are influenced by training. J Neurophysiol
36. Shim JK, Karol S, Hsu J, de Oliveira MA. Hand digit control in children: motor overflow in multi-finger pressing force vector space during maximum voluntary force production. Exp Brain Res
37. Sitburana O, Jankovic J. Focal hand dystonia, mirror dystonia and motor overflow. J Neurol Sci
38. Sohn YH, Jung HY, Kaelin-Lang A, Hallett M. Excitability of the ipsilateral motor cortex during phasic voluntary hand movement. Exp Brain Res
39. Vallbo AB, Wessberg J. Organization of motor output in slow finger movements in man. J Physiol
40. Walsh EG. Physiological finger tremor in medical student and others. In: Findley LJ, Koller WC, editors. Handbook of Tremor Disorders
. New York (NY): Marcel Dekker Inc.; 1995. p. 63-102.