You are on page 1of 13

IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING, VOL. 25, NO.

4, APRIL 2017 357

Proactive Selective Inhibition Targeted at
the Neck Muscles: This Proximal Constraint
Facilitates Learning and Regulates Global Control
Ian D. Loram, Member, IEEE, Brian Bate, Pete Harding, Ryan Cunningham, Member, IEEE,
and Alison Loram

Abstract — While individual muscle function is known, control usually arises later in life. For example, individu-
the sensory and motor value of muscles within the whole- als depending professionally upon skilled motor performance
body sensorimotor network is complicated. Specifically, the acquired through years of training may experience difficulties
relationship between neck muscle action and distal
muscle synergies is unknown. This work demonstrates jeopardizing their livelihood. Deterioration of motor skill and
a causal relationship between regulation of the neck problems of coordination, pain and injury can accumulate to
muscles and global motor control. Studying violinists per- chronic conditions in a variety of contexts. The nervous system
forming unskilled and skilled manual tasks, we provided uses immediate reward to drive reinforcement learning [1].
ultrasound feedback of the neck muscles with instruction If departing from the established control pattern leaves one
to minimize neck muscle change during task performance
and observed the indirect effect on whole-body movement. unable to function, professionally or otherwise, the path to
Analysis of ultrasound, kinematic, electromyographic and re-learning is rewarded less than the current sensorimotor
electrodermal recordings showed that proactive inhibition control. When complex patterns of control are involuntary
targeted at neck muscles had an indirect global effect reduc- and depended upon, the extent to which those patterns are
ing the cost of movement, reducing complex involuntary, flexible and the appropriate method to facilitate learning are
task-irrelevant movement patterns and improving balance.
This effect was distinct from the effect of gaze alignment open questions.
which increased physiological cost and reduced laboratory- Our general hypothesis is that the segmental structure of
referenced movement. Neck muscle inhibition imposes a the body provides a basis to the organization of motor output
proximal constraint on the global motor plan, forcing a which can be exploited to facilitate sensorimotor learning. The
change in highly automated sensorimotor control. The prox- planning of distal motor patterns may depend upon constraints
imal location ensures global influence. The criterion, inhi-
bition of unnecessary action, ensures reduced cost while applied at proximal locations within a kinematic chain. Adding
facilitating task-relevant variation. This mechanism regu- a proximal constraint to the task goal should change the control
lates global motor function and facilitates reinforcement of all distal segments. To test this hypothesis we chose a
learning to change engrained, maladapted sensorimotor proximal node likely to show a strong effect on distal motor
control associated with chronic pain, injury and perfor- patterns. Almost all sensorimotor tasks (looking, upright bal-
mance limitation.
ance, locomotion, reaching and grasping) require control of the
Index Terms — Neck, rehabilitation, sensorimotor control. head relative to the trunk. For reasons of proprioception, axial-
I. I NTRODUCTION appendicular and proximal-distal neuromuscular organization,
the neck is likely to influence processes planning motor output.
T HIS investigation arose from the practical problem of
changing highly engrained sensorimotor control in skilled
musicians. Motivation to change established sensorimotor
Using visual ultrasound feedback, participants targeted proac-
tive inhibition at the neck muscles to alter neck muscle action
during manual tasks. To achieve a desired goal, consistent with
Manuscript received April 30, 2016; revised September 24, 2016; neck muscle inhibition, should require changes in the global
accepted December 7, 2016. Date of publication December 21, 2016;
date of current version April 11, 2017. motor plan forcing a change in highly automated sensorimotor
I. D. Loram, B. Bate, P. Harding, and R. Cunningham are with Cognitive control. Constraints applied proximally may interfere with
Motor Function Group, Manchester Metropolitan University, M1 5GD, or facilitate the task. We test the idea that minimization of
Manchester, U.K. (e-mail: i.loram@mmu.ac.uk; b.bate@mmu.ac.uk;
p.harding@mmu.ac.uk; Ryan.Cunningham@mmu.ac.uk). proximal action unnecessary to the task, may make the task
A. Loram is with the Birmingham Conservatoire of Music, easier. We state three specific hypotheses.
Birmingham City University, B3 3HG, Birmingham, U.K. (e-mail: H1: Neck muscle movement can be regulated voluntarily
alison@loram8.freeserve.co.uk).
This paper has supplementary downloadable material available at while maintaining task performance.
http://ieeexplore.ieee.org, provided by the authors. This downloadable H2: There is a causal relationship between voluntary regula-
zip file includes the document “Supplementary material on use of tion of neck muscles and global control of movement.
ultrasound for biofeedback and analysis Final Version.pdf” and a folder
“Supplementary Material Graphic and Media Files” containing multime- H3: Proactive-selective inhibition targeted at the neck mus-
dia material referenced in the document “Supplementary Material.pdf.” cles reduces the global cost of movement.
The document includes the following sections Supplementary methods Selective inhibition is the ability to prevent muscular action
on this use of ultrasound for biofeedback and analysis, Supplementary
results, and Supplementary Discussion. incongruent with the task goal, while concurrently allowing
Digital Object Identifier 10.1109/TNSRE.2016.2641024 functionally relevant muscle action [2], [3]. While some
1534-4320 © 2016 IEEE. Translations and content mining are permitted for academic research only.
Personal use is also permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

Conflict between focus of attention controlled for the effect of wearing the US probe and famil- theory and our results is considered in discussion. in addition to normal task and neck regulation is considered to promote. Ultrasound (US) feedback of the movement within five dorsal layers. During inhibitory neck muscle regulation from gaze alignment. Combined with the task goal. but the key principle is that the nervous system organizes control of the whole body in relation to the task goal. mediated neck feedback. [17]. spinalis capitis. Experimental Design: Discriminating effect of voluntary reference or stabilizing segments to distal segments. Gener- ally. Also there is evidence linking head orientation and head con- trol instruction to global postural control [6]. may enable us to directly alter neck muscle behavior. foot) that is defined by the task goal [11]. which is argued to constrain of visually aligning the gaze to an external target. Bilateral seg- regulation of global control and facilitate learning. Attention (referred to as “neck regulation”). wearing the US probe and familiarization with the US information. 1. intrinsics) at 1 s focus of attention and conscious control of movement both intervals (red through orange to yellow. VOL. Transverse. Hence. APRIL 2017 selection occurs at the levels of spinal cord and brain stem. 4. from that or inhibit detrimentally automatic control mechanisms (the of using proactive. whereas an external focus behavior. head. Series D: US familiarization. selective inhibition to regulate neck muscle “constrained action hypothesis”). [13]. bilateral ultrasound analysis of neck muscle to voluntary regulation. diverting resources from accomplishment of the activation patterns. Tasks were carried out in five experimental conditions designed to separate the reaching movements activation can proceed temporally from effects of gaze alignment and US neck feedback from the effects of the head-trunk to end of the arm [8]–[10]. and reflexive their gaze to a blank US monitor. activation of muscles proceeds sequentially from proximal Fig. Series even when global activation is more synchronous. main goal and at worst detrimental. and 4 s) increase cost and impair task performance [14]–[16]. This phenomenon is known as anticipatory postural adjustment. stabilizing segment. To assess the indirect. 25. introducing conflicting Since US feedback of the neck muscles is given visually. finger. Fig. In support of our hypotheses. 2. 1 and 2). 2. This study uses ultrasound to provide feedback of on the environment (“external focus of attention”) and feed. Series B: Using the laboratory violin with an US probe attached to the participant’s neck. overlaid showing five layers of muscles Contrary to our hypotheses there is evidence that an internal (trapezius. the most powerful mechanisms for selection and modification of selection (reinforcement learning) are centralized and occur through the slow frontal-striatal loops [2]–[4]). and reports observation to secondary goals unrelated to the main task goal is at best of the indirect effect on whole body movement and muscle distracting. the neck muscles are amenable differentiated. we used a condition in which participants aligned mode of control by utilizing unconscious. 1. More generally. fast. 0. 3 shows our a dorsal cervical location with transverse probe orientation. It is unclear whether. Good control requires close reg. the planning of global motor patterns may depend upon constraints applied at the neck. . neck muscles during task performance provides a method that uninstructed effects we measured whole body movement. The difference between control processes [15]. Hence. Series C: of the proximal neck–trunk–legs anticipates and is typically Aligning the gaze. attention to the movement of one’s own body (“internal US provides a bilateral view of five layers of muscles focus of attention”) and feedback about the actual execution of (Figs. [7].358 IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING. Series E: Using US organized around movement of the distal end segment (eye. NO. a more automatic and more efficient conditions. Understanding this organization may help to promote Fig. An internal focus of attention is thought to encourage the experiment required control conditions to isolate the effect conscious control of movement. spinalis cervicis. Proactive refers to the use of environmental information to prepare the forth- coming inhibitory action [3]. laterally inverted US view of dorsal neck at vertebral level C3-C4 as background image (first frame). activation A: Normal task performance with no US. The effect of any constraints applied proximally accumulates along the kinematic chain from the reference. Visible muscle movement includes active a movement (“knowledge of performance”) is considered to be contraction and changes caused passively by action of other less beneficial than attention to the effect of one’s movements muscles. to the distal end segment. iarization with the visual US information. the current configuration of the body provides input to the construction of global motor output [12]. subject to the constraints of To assess the direct effect on the neck muscles we used performing designated tasks. 3. Further conditions of proximal elements [18]. activity. these latter conditions reveals the exclusive effect of regulating ulation of the distal end goal while allowing higher variability the neck muscles during task performance. Using muscle activity and skin conductance. there are known differences in innervation and control of axial and appendicular muscles [5]. the neck muscles with instruction to reduce any unnecessary back about the extent to which a movement accomplished the change of the neck muscles during task performance intended goal (“knowledge of results”) [14]–[16]. hypothesis testing flowchart. In some stud- ies. mentation after the experiment. splenius.

undermining a random order design. We test H1. the indirect effect on psychophysiological cost using their own instrument. All subsequent muscle movement whilst performing. Participants used their own shoulder rest and Series E: US mediated inhibitory neck regulation. and play part of the second study by Rodolphe Kreutzer”. participants references to “violin” include both violin and viola. 12 female) who participated in the To test the effect of observing and formulating the experiment. Following each task. The playing position was always maintained for more than 5 s. so this order A-E was constant for all Fig. H2. To test the A.: PROACTIVE SELECTIVE INHIBITION TARGETED AT THE NECK MUSCLES 359 progressed from relatively non-specific unskilled actions through characterization of the configurations sustained in playing the violin. Procedure view of their neck muscles. informed consent to these experiments. eight were amateurs. . a neutral position in which their arms were relaxed at their side. the gaze (and thus keeping the head relatively ticipants gave written. We characterize what perform the tasks as normally and as naturally as changes occur using univariate analysis. C). and D) in place (Fig. they were asked to using Analogue data (EMG. Task 1: Raising the arms(without the violin): “bring both arms up to a playing position”. Metropolitan University (MMU). we test for significant participants. Ethical Approval and Participants effect of a change of instrument and the addition of an US probe taped to their neck. C. Par. effects of visually-mediated neck regulation (E v. still). Method. approved by the Faculty of Science repeated “wearing” the US probe and using the and Engineering Research Ethics Committee. The fifth played the laboratory violin. the indirect. MMU. skin conductance). conformed to the standards Series C: Aligning the gaze. participants conducted a sequence of attempt to minimize change in the US image tasks intended to sample a range of manual activities that while they performed the tasks. 14. Series D and E had irreversible educational effects. Manchester laboratory violin. Participants performed these tasks on instruction. with the necessary control series (A. uninstructed effect on whole body movement using Kinematic data and H3. A) and neck regulation distinct from gaze alignment to the monitor (E v. in US image. 3. and association between task movements and changes three were students at a local music college. participants were asked to repeat all tasks as in Series C while observing the live US B. To test bow. 6). data. 1). B. Task 5: Playing a study: “raise the violin and bow. range. 47 ± Series D: Using US for movement . all tasks were These experiments. Series B: Using the laboratory violin with an US probe II. direct instructed effect of voluntary regulation on neck muscles using US To determine participants’ normal movements. the Series A: Normal task performance with no intervention. and After familiarization. 5 viola) (age 19–74. participants were asked to repeat all tasks as performed at the Cognitive Motor Function laboratory.LORAM et al. laboratory violin. Task 4: Playing a scale: “bring the violin and bow to the normal playing position and play the three-octave scale”. Following preparation for motion analysis and EMG participants were asked to describe all observed recording.7. 10 were professionals. Task 2: Raising the violin: “bring the violin to the normal sustained playing position” Task 3: Raising the violin and bow as if to play: “bring the violin and bow to the normal playing position as if to play”. but looking at the blank US monitor. Using multivariate analysis. Task 6: Playing own piece: “raise the violin and bow and play your chosen piece of music” This sequence of tasks was undertaken in each of five series (below) aimed at testing the effects of gaze alignment (C) and neck regulation (E) on normal playing. were asked to repeat all Series D tasks. mean ± SD. using 1-D statistical parametric mapping of univariate time series (Fig. in Series B. To test the effect of aligning set by the latest revision of the Declaration of Helsinki. Four viola players chose to use their own instrument for the indirect effect of minimizing change in neck all trials.visual familiarization. M ETHODS attached to the participant’s neck. Of the 21 players (16 violin. starting from and returning to. participants were given an extended period of changes in US image and specify when in the familiarization playing their own violin and an instrumented task those changes occurred. and when those changes occur possible.

To capture muscle movement not represented by the standing posture. US. splenius. NO. thighs. 14 surface (wireless) EMG electrodes and the total angular momentum of all segments around the (Trigno. rotatores) were automatically segmented at seven key movement. Due to technical issues individual parameters. we also feet. 50 mm field of view Aloka ProSound-5000). respectively. points around the E. and 19 participants. pelvis. and 9 sion. US quantities mea- occurrence at any particular instant. This method. neck. violin strain (Cardiacare. custom-made wireless device [19]. an “arm-raising event” was defined when the D. Using Visual 3D (C-Motion). Kinematic and Analogue Data left finger marker first raised 20 cm above the origin of the pelvic segment (time zero. Violin strain gauge. trials. Fig 4). APRIL 2017 We encourage the reader to access our supplementary mate. each with three degrees A 10-camera motion analysis system (VICON. chinrest force) measure physiological cost (14 muscles ×2 measures. three participants were excluded from Multivariate analysis: US. A custom-built. Fig 4). Herongate) were placed on the palmar aspect of gauge and skin conductance were normalized to a robust the second and fourth finger of the right hand for differential estimate of the maximal value (90th percentile) from all their recording. The splenius. spinalis they occur. 25. left and right muscles from five layers (trapezius. low-pass-filtered EMG data. Analogue quantities skin conductance. Univariate analysis: For each dataset. (RAISED minus PRE) were analyzed in separate datasets . 2). Kinematic quantities mea- and skin conductance data were collected successfully for sure whole body movement (H2). In lab coordinates. Processing of US. and 1000 Hz. of whole body movement. Apparatus and Measurements neck (C7) and atlanto-occipital (AO)). All upper and lower trapezius. This panel the muscle movement in a form that has a clear geometric shows common patterns between series for all variables. shanks. Participant was a random factor. upper arms.f. alis and medialis. rotatores / multifidus (Fig. 5. provide maximally a 30 parameter state of physiological cost. segment rotation. fits US. whole body movement and physiological cost for frames (2 s intervals). We subject the change a generalized neck segmentation model. These six segmental measures capture all aspects of shows the quantities most significant for Series. 7. sured compression force applied to the standard chinrest of the accumulated positive change (cEMG) where the laboratory violin. from the intervals −2 to −1 s and 1 to 8 s (c. respectively. for each frame. Kinematic and Analogue states analysis giving an US dataset of 18 participants. The meaning. in data collection. Statistical Analysis boundary of each segment were tracked automatically for. Panel D in Figs. For each trial. an eighteen-segment. clavicles. C. reported elsewhere.360 IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING. measures ANOVA with Greenhouse-Geisser adjustment for the median tracked boundary points were used to calculate deviations from compound symmetry. 14. rial on this use of ultrasound for biofeedback and analysis kinematic model was fitted to the marker data using “six- including example videos (Supplementary Methods). VOL. spinalis capitis. a T-shaped US probe  n E M G i > 0 E M G i (linear. wrists. US gives a 60 parameter state (10 muscles x 6 reader is encouraged to focus on those patterns rather than measures) of neck muscle movement. kinematic and analogue datasets. Oxford of freedom (extension. gastrocnemius and semimembranosus mus. EMG data were high-pass-filtered (10 Hz) to remove offset. knees. thorax. For each dataset (US. were Metrics) and whole body marker set was used to track calculated. Galvanic skin conductance (GSC) was recorded via a rectified and then low pass-filtered with a cut-off at 5 Hz. seventeen model-based joint-rotations (ankles. spinalis cervicis. Within the trans- (RAISED). 4. 1). and Task (six levels 1–6) were within subject factors and centroid cumulative absolute movement and cumulative expan. and sustained playing position shape (Fig 2 and Supplementary Material). cles. 12. wireless strain gauge device mea. segment area. Having shaved calculated the whole body center of mass location (CoM) and cleaned the skin. Adhesive gel electrodes For each participant. to provide a transverse This measure counts all prior increments irrespective of when view of five muscular layers (trapezius. to give a 51 parameter state (17 joints x 3 dof) 18 segments (head. hips. Analogue quantities (EMG. For Series B to E. tibialis anterior. skin conductance. cE M G n = if was attached to the dorsal neck of each participant at the level i=1 other wi se 0 of C4 using micropore tape (Fig. elbows. intrinsics (multi- change in state (RAISED minus PRE) represents neck muscle fidus. adduction and axial rotation). strain gauge) and effort (H3). For muscle EMG we calculated an additional measure. 21 participants were included in this analysis.5 MHz. degree of freedom” inverse kinematics. Series (five levels A–E) the centroid location (x. all parameters wards and backwards from each key-frame using a KLT (RAISED minus PRE) were tested individually using repeated tracker [21]. forearms. sterno-clavicular. Delsys) were placed bilaterally to record data from CoM (Visual3D function model_angular_momentum). 100 Hz. 7. analogue) we calculate the change between initial analysis—segmentation of muscles and tracking of muscle neutral standing state (PRE). shoulders. ferent prior times are given equal weight at the instant of kinematic and analogue data were synchronously sampled at comparison rather than being diluted by a low probability of 25 Hz. hand. vastus later. These states were calculated using mean values verse plane. y). US video sequences were subjected to two stages of kinematic. thoraco-pelvic. From this model. EMG sure neck muscle movement (H1). spinalis cervicis. For each segmented muscle. Intermittent bursts in activity occurring at dif- capitis. independent MRI-US dataset of 24 participants in the normal upright standing posture [20]. derived from an in state to univariate and multivariate statistical analysis.

with show the trial scores from the first two of these lin. is recorded the gaze alignment trials (Series C) were not. discriminant functions (DF1. p <0. panels B/C in Figs.: PROACTIVE SELECTIVE INHIBITION TARGETED AT THE NECK MUSCLES 361 Fig. whereas s) is recorded as RAISED.trapezius. 0. Correlation of DF2 with all US quantities showed between Series in canonical units of covariance. with Series C. DF1) to increase muscle movement. no quantities showed these eigenvectors provide patterns of neck muscle movement. Correlation of all US for statistical analysis. DF2 accounted for 28% of the variance between series.splenius.9. These trial ear. 4 s). grouped by series. C: Neck regulation and gaze alignment were associated with with the arms hanging vertically. showed a tendency (absolute. different from Series B (blue). SpiCap-spinalis mean Mahalanobis distance between centers.0001).. of monotonic (DF1) and bitonic (DF2) patterns through Series A to E. Distance Series E from C was 0. For all quantities ANOVA. and 9 using univariate repeated measures ANOVA. Direct instructed effect of neck regulation on neck muscles (US) A: Mahalanobis distance between series. 3. the participant raised the violin using the distinct patterns of effect on neck muscle movement. Spl-splenius . canonical. B. but also partially common to US familiarization and neck meters and 18. (iv) vertical height of third metacarpal marker showed a significant difference between series (repeated measures relative to origin of pelvis segment (LFin wrt Pelvis z). for individual quantities. significant interaction between factors Series and Task. 7. Rows in descending order (i) Centroid x-displacement in transverse standard deviation. and relative to the monotonic pattern of For US.1. distance is a scale-invariant. Horizontal left arm. Representative task and measurements Starting from neutral B. Neck regulation had a greater effect on neck muscle movement than gaze alignment. muscle movement. DF1 from left upper trapezius. 7. To reduce the multiple variables to orthogonal univariate quantities which maximize the separation between Series. 21. and the regulation (Series E) and US familiarization trials (Series D) were mean value during time 1 to 8 s (or when LFin wrt Pelvis z less than 0. 2.2 significantly different from normal performance (Series B). the superposition calculated and ranked in descending order the four eigenvec. Bars also shows effect size.8. represents a systematic reduction in muscle movement generally for all muscles and distinct pattern of variables.spinalis cervicis. Raising the left hand 20 cm above the pelvis defined an event axis shows trial scores ±95% confidence intervals from the first two and time 0 s. regulation. and 9 the show means ±95% confidence intervals. Fig. 4. DF2) in units of A. red. Int-intrinsics . The change. DF2). This distance D: The most significant univariate effects of Series on neck muscles (US) in descending order of significance rightwards to alpha<0. Red indicates series which differ significantly US image of left neck muscle segments (blue. Quantity horizontal link joining Series or clusters of Series shows the labels are Trap-trapezius . and represented an orthogonal pattern most strongly associated with gaze using respectively 60. 5. capitis . as above.0001). Tukey-Kramer post hoc pairwise comparisons.005. 5. D and E all significantly different from Series B. magenta. whereas gaze alignment (Series C) had a distance of 1. 14 participants.05. At p < 0. Visual feedback with instruction to minimize unnecessary neck tors of the matrix (between series sum of squares / within muscle movement (Series E. Relative to normal task performance (Series B). orthogonal canonical discriminant functions (DF1.LORAM et al. green-intrinsics). neck regulation (Series E) had a Mahalanobis distance of 1. yellow. . and muscle cost discriminating Series A-E. 5. SpiCerv-spinalis cervicis . 1. Mahalanobis DF2 represented mainly an increase in muscle movement. most significantly for the left neck muscles. Gaze alignment (Series C). exp-cumulative expansion. alignment. Kinematic and Analogue quantities respectively. I-cumulative absolute movement.spinalis capitis. 29 (strain gauge excluded) para. For Panel A in Figs. whole body movement. A-area. These scores were also tested for correlation with all quantities to scores were tested for significant difference between series produce the “structure matrix [22]. To show these patterns. Tukey-Kramer pairwise comparison showed neck the mean value during times −2 to −1 s is recorded as PRE. DF1 accounting for 54% of the variance between series. resulted in a consistent series sum of square) [22]. Within Matlab 2015b. B Kinematic analysis fitting 18 segment “6 DoF” quantities with DF1 showed DF1 represents a general reduction in kinematic model to Vicon marker data at times (−1. C. we Univariate analysis confirms. 51. Each eigenvector. and unassociated with gaze alignment (Series C). multidimensional measure of separation between points in units of variance. RAISED minus PRE. DF2 showed a significant difference between series (p We calculated Mahalanobis distance to quantify distance <0. and D partially). (ii) left upper trapezius (LUT) accumulated positive change EMG (solid) represented a pattern most strongly associated with neck regulation and accumulated absolute movement US (dashed) (iii) surface EMG (Series E).

distinct from all chosen statistic at each temporal sample and calculates a series delays and reduces whereas gaze alignment increases turning the head towards the violin. 6CDEFH. Fig. participants completed all tasks suc. 6B).362 IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING. all holding the violin conventionally on the left show the cumulative step-like change in muscle movement and side between shoulder and chin. most salient in the left spinalis capitis and For all Series A-E. p =0. respectively (Fig. To determine when the indirect effect of Material (SM)). Subsequent scoring of of the entire data set. under the chin (approx. the false positives of multiple scalar tests and avoids the false negatives of scalar tests with Bonferroni correction. neck regulation (Series E) task and is less sustained. 6A-B. activity. During data collection play.org. up. NO. 1. Axes A: Accumulated absolute US movement of neck muscles tested. 6AB. “anova1rm”) [23]. we apply SPM1D to the temporal evolution of the position though there was an altered. Fig. raising the violin with the left eigenvectors (DF1. F: Vertical left hand movement. Using the flowchart of Fig. These trial Representative trial. 4. −0. 25. 3. effect of series) relative two experienced assessors each showed no significant effect of to the threshold of significance for the quantity tested. 4. using the kinematic Fig. and “ fitrm ” were used respectively head flexion. threshold of significance appropriate for the partial indepen. 3–6 Supplementary US data (Fig. VOL.9 s mean times for to the evolution through time of the trial scores of the first Series A-D. www. DF2). R ESULTS the neck muscles. placing and holding the violin show significant effect of Series at p < 0. SPM1D calculates the H: Atlanto-occipital axial rotation to left. consistent with regulation of position to a predetermined value. sound was produced later neck regulation on global movement occurs we apply SPM1D at 4. smoother trajectory to accumulated absolute movement of the neck muscles from final positioning during series E (Figs. Panels C-R show mean values for each series: Time zero is when the left finger first rises 20 cm above the pelvis origin. resulting in reduced lateral flexion of the head onto the violin. Time series analysis: to assess the timing of signifi.0.05. shows consistency in space and time of cant difference between Series A-E. D. and 4 × 6 trials is simultaneous with and sustained throughout the task. 4 shows a participant. Shows effect during Series A to D. 3 the change . Fig. preserved initial joint angle. forwards. 4. Neck regulation delayed and reduced head movement to calculate the canonical discriminant functions and conduct whereas gaze alignment (Series C) increased head extension and then the repeated measures analysis. 1–2 s) (Series B.5 to 1 s). Shows stronger. all Series C-E eliminated the initial functions “ manova1 ”. US and EMG analysis of the left upper trapezius both cessfully. APRIL 2017 Fig. SM). Head Control: C: Left spinalis capitis accumulated US movement.1. B: Kinematic discriminant function scores (DF1 .M0. Both hands showed no signif- determine when and where the direct effect of neck regulation icant change in speed or timing during raising to the playing occurs. 6O-Q shows the temporal evolution of the the randomized recordings (Tasks 4-6) for playing quality by F-statistic (repeated measures anova. reduces transient muscle movement whereas gaze alignment (Series C) adds movement. common to all series requiring gaze alignment (C. E) (DF1 ) anticipates D: Left spinalis capitis US x-displacement. 3 are right. cervicis. 2.44. Shows exclusive shows an effect of neck regulation distinct to Series E and not arising effect of neck regulation on global pattern of movement (E-C). selected univariate time raising and holding the violin during all series. Evolution through time of effect of neck regulation A-B temporal evolution of effect of Series (mainly neck regulation): z is SPM (F) relative to threshold of significance for quantity tested. DF2 ). All time series presented in Fig. Starting from the neutral position. SPM1D avoids regulation whereas Series C is consistent with focus to the US monitor. Execution is reflected by clear preceding and accompanying changes in III.3. neck regulation. 6A). two kinematic discriminant functions (Fig. We interpret Series E results as consistent with ongoing inhibitory neck dence of measurements repeated through time.81 resp). To Series (p =0. series were tested using 1-D statistical parametric mapping G: Neck right lateral flexion relative to thorax.7 s compared with 3.6. ing performance was perceived by the experimenters as Group results: All following results report group analysis indistinguishable between conditions. (DF2 ) naturally during the sequence of familiarization trials. Task 2). During Series E.spm1d. earlier effect on appropriate muscles for the movement sequence left neck flexion/head turn followed by right arm action. 3. neck regulation is distinct from all preceding series in delaying and reducing neck movement implemented by the open-source toolbox SPM1D (v. 6 hand (approx. E: Atlanto-occipital extension. scores were calculated at every sample.

unique to neck regu. with regulation of global balance and global rotation during the disturbing effect of the tasks. Shows neck regulation minimizes unnecessary over generation and distinct from all preceding Series reduced shoulder protraction toward the cancellation of upwards momentum when raising hands-violin to playing violin shoulder rest. Following the flow chart of and neck regulation on the whole body movement requires Fig. shows the CoM. c. 3). Fig. muscles. 6JKLMN. Fig. Series A does not reducing activity in the right posterior upper leg associated with support. 6C-D illustrates normal task movement. 7A). Positive=right.LORAM et al. and earlier segment angles with DF1 and DF2. shows that neck regulation produced uninstructed changes in movement throughout the body. N: Skin Conductance: Neck regulation reduced the delayed rise in skin conductance associated with task performance. global leftwards frontal rotation around K: cumulative positive change in left vastus lateralis EMG. Fig. Neck gaze alignment (C-E).B. sustained in playing position (2–8 s). 7B) and reducing neck muscle movement and produced a pattern of also a substantial pattern discriminating neck regulation from effect distinct from the effect of alignment of gaze. shoulder protraction requires action of multiple position (0–2 s). minimizes unnecessary. Compared with subsequently strongest on right trapezius. the time series of Fig. ScrollThumb. (Fig.f. unobserved in their series. J: L Shoulder anterior displacement relative to thorax. 7 reports the key findings. instructed effect of neck regulation on neck and neck but also in the arms. 7D). The direct. Positive is clockwise around right. that neck muscle combining the effects of all joint rotations. whole body. distinct effect of neck regulation (Series E). involves increased effort. Gradient = instantaneous Fig. whole body movement RAISED minus PRE was tested systematically. 5) showed than Gaze Alignment (Fig. 8. muscle is retractive elevation of the scapula. neck and reduced laboratory-referenced axial rotation of the . There was a substantial that neck regulation using visual feedback was effective in pattern of effect common to both conditions (Fig. Bottom row: SPM(F) scores relative to threshold. The main text Considering all rotation axes from seventeen joints. Localised cost. Neck regulation lation. (Fig. Positive lab axes 1. Q: Positive is CR around vertical axis. Fingerboard. lower trunk and legs. forwards. gaze alignment common to contrasts an early. Neck regulation minimizes reducing anticipatory activation in the leg. The effect of neck regulation distinct from the with a general reduction in neck muscle movement (Fig. 2. The action of this minimizes left movement of CoM during playing position (3–5 s). Fig. not just those movement can be regulated voluntarily while maintaining task rotations which show significant univariate differences (Fig. ScrollFinger) M: Right semimembranosus EMG. Shows neck regulation is associated with chronic injury in violinists. and laboratory-referenced flexion of the head and observed during gaze alignment (Series C). 8).: PROACTIVE SELECTIVE INHIBITION TARGETED AT THE NECK MUSCLES 363 Fig. These patterns can be seen regulation had a greater effect on neck muscle movement in the effect of Series on joint rotations throughout the body than gaze alignment (Fig. global angular momentum around CoM. 6OPQR. ing the violin on the left side of the body. shows an effect of neck regulation tracked the violin segment during Series B-E. distinct effect of neck regulation (Serie E) in reduc. 3. Skin conductance B.B. in activity associated with raising the left arm. Neck muscles regulation had a greater effect on whole body joint rotations Analysis of all US neck muscle quantities (Fig. neck regulation. Neck regulation was associated (Fig. shows a reduction. We report a summary combining the structure matrix of Fig. include Violin. costly. 6Q. N. leftwards global axial twisting sustained in playing position (1–9 s). R: Center of mass location. Figure captions report detailed results. For individual parameters. 3 are I: Thorax axial thoraco-pelvic rotation. on the left neck muscles. The indirect uninstructed effect of neck regulation on measures solely sympathetic arousal. L: Left upper trapezius EMG. increases grip force on the violin and P: Positive is CR around forwards horizontal axis. performance. relationship between voluntary regulation of neck muscles and tern which included increases in neck muscle movement global control of movement (Fig. 7C). Global balance and unnecessary cost: Panels associate neck inhibition. O: Positive is clockwise rotation (CR) around right horizontal axis. 5A). was strongest. Neck regulation increases rightwards rotation of thorax towards sustained playing position at 2s. N. (A-C) and (A-E) reduced atlanto-occipital (AO) and neck ing muscle movement and the increased muscle movement flexion. particularly left spinalis cervicis and and also provide the visualization of Fig. most significantly in the head A. Four markers (Chinrest. 5B) effect of gaze alignment confirms hypothesis H2 of a causal whereas gaze alignment was associated with a different pat. 5D shows these Complete representation of the effects of gaze alignment distinct patterns across Series. up. 6. O-Q show time integrated angular momentum which equates to rotation of whole-body mass around the global centre of mass (CoM). 5C). these results confirm hypothesis H1. vertical axis. 6 A shows how the effect of neck regulation occurred correlations of all joint rotations and laboratory referenced early during raising the left arm.

B). 9B. that In summary. neck regulation and gaze alignment Col 2 (Gaze alignment) shows the difference (A-C) and thus shows the were associated with distinct patterns of effect on global joint rotation joint rotation present normally (Series A) and removed by gaze alignment (RAISED minus PRE). C: Trial scores of first two discriminant functions (DF1. exclusively positive. skin conductance whole body above the right hip (Fig. neck regulation per se removed displacement of body parts Fig. downwards compression of the head relative to the Neck regulation was associated (DF1) with decreased cost thorax. NO. 9A). 6 O-R). 3 upwards (axial). DF1 (46% of variance). associated with reductions whereas gaze alignment was asso- ciated with increases in all muscle activities. analogue quantities. 25. For trials with recorded and C significantly different from Series A. for the most significant univariate effects of series on joint rotation in descending order of skin conductance and chin rest compression. Distance of Series E from C was also 1. 6C-H). 3). 8 cols 2. with only Series E the physiological cost of the task. shows movement removed by neck regulation which is not Fig. NkTho-Neck relative to thorax. 7.005. 7B) were exclusively positive. 8 and compressive force on the chinrest. DF2 joint rotation present normally (Series A and removed by neck feedback resp. contrast (C-E). whereas gaze alignment had a Col 1 (zero) shows the null case of zero difference (A-A) and thus mean distance of 1. (Series C). Following Fig. neck regulation (E-A) was a reduction in muscle activities. shows the mean pre task kinematic location (PRE) of all trials from all participants in Series A. regulation on head control (Fig. ShCla. For each series the mean joint rotations were computed (RAISED minus A: Mahalanobis distance between series. ing that the effect common to gaze alignment (C-A) and Task interaction was confined to thoraco-pelvic and right ankle joints. meaning that pure neck regulation was Upper arm relative to clavicle.). whereas indirectly reduces the global cost of movement. The superposition of patterns of DF1 and DF2 is seen for individual quantities. particularly the (Fig.0001) and significant (Series C) and removed by neck feedback (Series E). Significant Series. (Series E). E). Task-independent movement eliminated by gaze alignment whole body movement. these results confirm hypothesis H3.7 (Fig. violin strain gauge) each provide a measure of studied required participants to lift the violin with their left . There were significant univariate effects of Neck regulation Movement DF1 (Fig. DF1 showed (C-E) and thus shows the joint rotation present during gaze alignment a significant difference between series (p < 0. 6I-N) and global C. In common with the US analysis. a mean Mahalanobis distance of 1. represented a pattern common to all series Col 4 (Neck regulation minus Gaze alignment) shows the difference requiring gaze towards the US machine (Series C. 8. Rotation axes 1 right (extension).364 IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING. 7C) were relative to neck (Atlanto-occipital). Gaze alignment was associated with a left shoulder and abduction of the arms relative to the thorax. Wri-Hand relative to forearm.0001). Fig. Neck regulation had a PRE). Furthermore. forward movement of the shoulders. lateral and extensor rotations of the head and had a greater effect than gaze alignment (Fig. Neck regulation quantities. D and E from A. skin including left shoulder protraction (Fig. 6 details the timing and nature of the effect of neck relative to the trunk while redistributing axial rotation. difference of Series C. and neck. The indirect uninstructed effect of neck regulation on balance (Figs. 9C and D). and neck regulation. significant analogue correlations with D. ThoPe-Thorax relative to pelvis. DF2 (30% variance) represented a pattern discriminating Neck regulation (Series E) from Gaze Alignment (Series C). different pattern (DF2) of increased cost (Fig. all significance to alpha< 0. uninstructed effect of neck muscle regulation on Fig. arm and leg movements. 9 reports independent analysis of combined analogue removed by gaze alignment. Neck regulation had the greatest effect on cost alone reduced axial. 4. Relative to normal (magnification x5) added to the mean pre task kinematic location (PRE) task performance (Series A. Red indicates significant different from Series A (blue). 2 forwards (lateral). AO-Head significant analogue correlations with DF2 (Fig. Indirect. 6A). DF2 showed significant difference between series (p < 0.9. Bars show means ±95% confidence intervals. Cols 1–3 show the difference in mean joint rotation between series greater effect on Movement than Gaze Alignment. and D). and increased axial rotation of the thorax relative to the pelvis. Col 3 (Neck regulation) shows the difference (A-E) and thus shows the B. 6J-N). localized cost within the lower back. APRIL 2017 Fig. neck regulation (Series E) had of all trials from all participants in Series A. D. The tasks conductance. Neck regulation (Series E) reduced analogue quantities anticipatory and sustained activity in the legs and upper body The analogue quantities (fourteen muscle activities. upper limb and legs (Fig. 3.9. VOL. mean- (Series E-A) on neck head. gaze alignment removed movement of the head proactive-selective inhibition targeted at the neck muscles and trunk relative to the laboratory frame of reference.

to DF2.LORAM et al. securing the instrument and playing ment had some. Neck regulation had the ment increases gravitational moments [24]. whereas gaze neck regulation improved global balance. skin conductance and compressive based upon their own observations. D). DF2 (23% variance) was most strongly associated with gaze alignment (Series C). Red indicates significant while maintaining task performance difference from Series A (blue). (movement-visual association). but less. D. effect of neck muscle regulation on cost within the frontal plane (AR) (Fig. reducing global cost. turn the head leftwards onto the chin rest to support the weight of the violin between chin and shoulder rest. through internal effort transmitted externally through the forceplate. H2: There is a causal relationship between voluntary regu. However. Violin–violin changes in whole-body movement. the implication is that observation and formulation of the image with different phases of the task such as raising associations between self-movement and neck muscle move- arms. Both neck regulation and gaze align. 6R). 6O). performance of unskilled and skilled manual tasks. and C all significantly different from Series muscles and the involuntary construction of whole body motor A. summarized below. to investi. GM-gastrocnemius medialis . SM-semimembranosus . 6R).0001) with Series E. D: Significant univariate effects of series. and sustained neck muscle movement. the procedure of Series E was not revealed during Series participants in Series D had to see and associate changes in D. TA-tibialis anterior . task actions can perturb global balance including location of whole body center of mass (CoM) relative to base of support and rotation of whole-body mass about the CoM. using force on the violin chinrest. visual feedback for guidance. Next. labels are UT-Upper trapezius . 5 and 6). A: Mahalanobis distance between series. regulation was distinct from the effect of gaze alignment Bars show means ±95% confidence intervals. Left CoM displace. increased skin conductance and geted at the neck muscles (“neck regulation”) (Series E). are ranked left to right with most significant left.0001). neck regulation IV. and with improved global balance. and reduced anticlockwise global mass-rotation around CoM Fig. turning the head. is a deviation from the overall progressive change through Series A-E maximized in Series C This DF2 pattern shows gaze alignment to cles reduces the global cost of movement be associated with an increased transient activity (accumulated positive Using visual feedback. reduced sustained left CoM displacement (Fig.05. Quantities significant at alpha< 0. with decreased physiological tions given ensured participants made their own associations cost in muscle activities. The instruc- achieved through gaze alignment.1. DF2 showed a significant difference between three hypotheses: series (p < 0.2. Corresponding H3: Proactive-selective inhibition targeted at the neck mus.9. selec. proactive-selective inhibition tar. These results confirm the possibility of voluntary. reduced laboratory-referenced movement of the head and trunk at increased physiological cost (Figs. in Series E. The main results. Acquiring movement-visual associations (Series D) caused tive inhibition of neck muscles without detriment to task some reduction in neck muscle movement (Fig.05. The distance of Series E from C is 1. symmetric orientation increases gravitational moments. Quantity (Figs. DF1 (55% variance) was most strongly associated with neck regulation (Series E). In these experiments. participants had to make the necessary changes in motor command that simultaneously reduced unnecessary neck muscle movement and achieved task A. proactive-selective inhibition targeted at the neck mus. Neck regulation reduced unnecessary transient upwards rotation (Fig. At p < 0. B. rotate the violin to rest on the shoulder. AR from the initial greatest effect on cost and had a greater effect than gaze alignment. VL-vastus ment resulted indirectly in uninstructed. with only Series C and B significantly different H1: Neck muscle movement can be regulated voluntarily from Series A. Note that participants were cles (Series E) resulted in changes in global movement not not instructed in the features of the US image. The effect of neck there were no significant interactions between series and task. Since performance. output. effect on neck muscle action. bring the bow to the violin and bow with the right hand to play. Skin–skin conductance. B) neck regulation (Series E) had a mean Mahalanobis distance of 1. increased chin rest compression. 9. confirmed our C. Gaze alignment (Series C) chinrest strain gauge. global angular and linear momen- tum are unchanged by segmental motion. Corresponding to DF1 is a lation of neck muscles and global control of movement progressive reduction in all quantities from Series A to E. D ISCUSSION and gaze alignment were associated with distinct effects on analogue This study used visual feedback of the neck muscles during quantities. In common with the US and movement analyses. 5B. 6–9). change) in arm and leg muscles. alignment had a distance of 1. showed significant difference between series (p gate the relationship between voluntary regulation of the neck < 0. Hence Relative to normal task performance (Series A. The structure matrix confirms global reductions in transient (cEMG) and sustained (EMG) muscle activity achieved clear general reductions in anticipatory. 6O-Q) and to lateral balance (Fig. 6P). task-independent lateralis . Trial scores from DF1 and DF2. The Voluntary Inhibitory Regulation of Neck Muscles performance (proactive-selective inhibition). Also . Indirect. transient associated with DF1 and increases associated with DF2. Neck regulation was notable in minimizing the disturbing effect to global mass-rotation around vertical and forward axes through CoM (Fig.: PROACTIVE SELECTIVE INHIBITION TARGETED AT THE NECK MUSCLES 365 hand. using visual feedback. By con- trast. In the absence of external forces and torques.

366 IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING. Hence. the leg muscles was reduced. in to many whole body movements. Neck Muscles Influences Construction of Whole formed the tasks. 6 –9). infer that neck regulation regulated performance by reducing the physiological cost. voluntary. Confirmation that participants engaged cle and movement. refined by experience. immediate reward provides little most movements including looking. inhibitory Instruction to minimize neck muscle movement provides a neck regulation. of which participants were largely unaware. unnecessary movement throughout monitor. Changes in whole body include changes in active contraction of the neck muscles and performance. participants were instructed to Body Movement minimize neck muscle movement on the US monitor while During Series E. C. synergistic contributions say that proactive-selective inhibition of the neck muscles from the neck muscles observed by US. before leaving the laboratory. B) and Series E was regulated by attending changes in movement require altered action from multiple to and minimizing neck muscle movement. Disturbance to global balance The natural. Sensorimotor control by trial. 6C. not just biomechanics. trunk muscle movement (Series E). We developed methods of ultrasound analysis [20]. throughout the body (legs. building upon the recently Series E reduced the task related rise in skin conductance. neck) in a process distinct from previous Series A-D is provided early in the arm raise (Fig. The neck muscles viewed using US include Vicon motion system recordings of their movement. Neck muscles change if the head moves or previously unobserved in numerous trials during familiariza. Forward protrac- (Fig. 6). VOL. 1 and 2 Supplementary Results). were passive shape change caused by the action of other muscles. Clearly there is potential iological input anticipating future demand for effort [26]. increasing axial rotation of the The second benefit is one of reducing cost and regulating trunk relative to the pelvis and laboratory may minimize axial performance. 5). The generation of new muscle movement (active or passive) will minimize whole behavior was uncomplicated. B) and also when minimizing neck changes we observed including shoulder protraction. in general. for developing augmented feedback. if the arms move. incentive to learn. splenius. 5D. reaching and grasping. 25. B. 4. Fig. and arms. neck rotation to maintain alignment of gaze with the US rized as reducing effortful. but not limited to. the monitor to reduce neck muscle movement (proactive. We infer that sensorimotor by the increased duration to produce sound in Tasks 4-6. trunk. participants were shown neck muscles. and were thus self-chosen For example changes in head rotation relative to the trunk by the participants. minimisation of neck tion and control series A-D (Fig. regulation of the neck muscles Sympathetic arousal reflects central integration of psychophys- would be possible without feedback. The Mechanism by Which Direct Regulation of the Participants were never told to change the way they per. The global playing (Series A. NO. Forward protraction and elevation of the right shoulder was selective inhibition) required coupling of observation with reduced. the central selection is altered to reduce neck muscle that is highly automated through reinforcement learning and movement change observed using US. We can therefore muscles including. previously unknown and unplanned. relatively instantaneous and self. body movement although exceptions to this generalization chosen (unsupervised). Anticipatory and transient task-related activity in motor command. Furthermore. We observed that. by coordinating change in all muscles to alter neck muscle Furthermore. Since. participants demonstrated execution of performance the neck muscles. When root of kinematic chains linking the head. altering control of the whole body to allow smoother performance with improved global balance. (Figs. criterion to alter the central selection of motor output. are possible. performance is important. movement. Most upper trapezius. APRIL 2017 focus of attention to the blank monitor did not replicate the the body (Figs. . control is altered. The change between normal rotation. Trial The first benefit is one of learning. through the procedure of balance and locomotion will produce changes observable in Series E. This result is achieved training is habitual and inaccessible to modification [25]. leg activations and chin rest compression. The effect of neck regulation can be summa. was reduced. spinalis cervicis and expressed surprise upon seeing their movement patterns in the intrinsics. effectiveness of regulation of the neck muscles (CoM location) and global angular momentum by the task observed in Series E raises the possibility that following train. Changes in neck though the mechanical speed of hand movement was unaltered muscle observed using US. Skin conductance measures sympathetic arousal. For example. ing with feedback. the automated solution. muscles crossing the neck respond typically lies within a multi-dimensional local optimum. Use of the information presented on tion of the left shoulder towards the violin was reduced. Downward flexion of the head and reduction in neck muscle movement observed during Series E compressive chinrest force was reduced. These muscles alone cannot explain the global normal playing (Series A. we observed simultaneous effects in mus- performing the task. nism associating neck regulation with whole body movement. Involuntary Effect of Inhibitory Improved performance at reduced cost provides its own reward Regulation of Neck Muscles on Whole for subsequent reinforcement. D). We observed two indirect benefits to voluntary. spinalis capitis. (neck regulation) caused the involuntary change in whole body Trial and error feedback provides one possible mecha- movement. arms. The Indirect. referred to as muscle “movement”. After completion of the entire experimental would cause passive changes potentially observable in all series. Body Movement We emphasise that no instructions were given to par- ticipants regarding how to change their task performance. the neck lies at the which local departures usually worsen performance. In Series E.

The mechanical rationale for the dependence of control During Series C. this literature is that automated control is superior both in tory neck muscle action and head movement would predict performance effectiveness and in cost of performance. dance. we propose facilitating task relevant variation. most predictive head-referenced estimate of body assumption that conscious control adds constraints detrimental location justifying the high biological investment in the abun. [28]. this lack of interaction implies the effect of Series experience of the experiment. Since task order was consistent. Normally. This straints in position and force defining the control of segments follows the location of the neck proximal to outgoing motor accumulate along the kinematic chain from the stabilising kinematic chains and incoming sensory chains of the head. The maladaptive nature of learned. or within a closed chain. Reconciliation With Theory of Focus of Attention quently specifying reduced co-activation across the neck joint There is evidence that an internal focus induces a con- would predict reduced compression of the violin between chin scious type of control. we suggest from personal randomized. Series and Task. Conse. external pre- erence frame. our results favour reward [1].. the feedforward mechanism in which planned neck action nism. detrimental to established performance.LORAM et al. future state of informs the global plan. neck and head movement would simplify sistent with the end goal. vestibular and straint of minimizing movement in the neck muscles. Along the kinematic chain task specific injury [30]–[32]. Our interpretation is that participants were the extended proprioceptive-kinematic chain [29]. For example. automated . has consequences for the planning and imposes a proximal. ensured acceptable axial rotation. The criterion. This result lessens support for of attention on the end goal. trunk location from the head. allow- no interaction between the two factors. Con. Our proposed mechanisms only exceptions were thoraco-pelvic rotation and right ankle suggest the criterion simplified planning. the head provides a second sensory ref. frame from which to estimate configuration. to performance. economical motor control Series C support the idea that the addition of constraints was requires a known reference. ensures reduced cost while triceps. Accurate. bi-articular muscles cross The proximal location ensures global influence. allowed increased external focus did not change through time. raising. reduced co-activation specified at the neck would require reduced co-activation recursively down an open chain. [15]. also favours the plan. motor system by interfering with automatic control processes For example. Minimization to a relatively fixed set of proximal segmental movements con- of anticipatory. securing and playing the whereas an external focus allows automatic control processes violin requires simultaneous control of all body segments to to regulate movement [14]. forcing a change in highly automated sensorimotor control. A reduced anticipatory control of the legs and trunk as observed second assumption is that conscious control adds constraints during the interval −1 to 1 s in Series E (Fig. [14]. density and elaborate structure of neck muscle spindles During Series E. US monitor provided augmented feedback i. The ing consideration of new solutions.. informs construction of the global motor regulation produced almost instant results. Automation had restricted their behavior at the neck influences every dependent estimate. participants aligned gaze at the blank US upon the neck is mirrored by the biological design of the monitor while performing the instructed tasks. segments. Proprioception is essential for extracting body sentation of intrinsic feedback. control of segments distal to the neck. participants consciously applied the con- ground provides a reference.e. The control of child segments depends upon In summary. extensor digitorum. sensation [12]. These results undermine the proceeding from the head. The reference provides a sensory detrimental since the cost of performance increased. The result was generation motion from head-referenced visual. The movement of any segment occurs relative to its feedforward mechanism. One assumption within ensure dynamic balance. rather than solutions were less effortful and. many global muscle activation/movement pat- the current state and planned futures states of the parent. The auditory sensation. Results from human sensory system. inhibition of unnecessary action. all joints from the head/neck to the finger (e. Instruction to minimize unnecessary planning the control of dependent. one in which the current and planned.: PROACTIVE SELECTIVE INHIBITION TARGETED AT THE NECK MUSCLES 367 Feed-forward processes provide a second possible mecha. Since the unaware of the constraints self-imposed by their learned sen- neck lies at the root of this chain. positioning. [15]. proprioceptive inaccuracy sorimotor mappings. proprioception of the neck provides assumption that automated control is superior and also the the first. constraint on the global motor plan. D. vestibular and auditory of new behavior in proximal components of the movement. as this explanation While involuntary automated control is learned predicts the effect of Series would increase through time. causing individuals to constrain their and shoulder rest. Selective inhibition of the neck muscles applied at the neck. reference segment to the distal end segment. the During Series E. Any specification upper limbs and trunk. the inhibitory nature and proximal target [27].g. Through visual. trapezius. [17]. terns are associated with changes in neck muscle state. the trial and error feedback mechanism. A more differentiated estimate of configuration of this conscious control removed automated constraints lim- accumulates through proprioception of additional joints along iting performance. Proactive minimisation of anticipa. flexor digitorum). since. Our informal observation that neck the neck muscles. The main mass of the body lies close to without detriment to task performance with lowered immediate the trunk and the neck is the primary articulation defining cost and with reduced movement associated with chronic. sensorimotor learning works by immediate Hence while both mechanisms are possible. neck muscle movement required inhibition of all automated Almost all quantities (US. kinematic and analogue) showed solutions causing unnecessary neck muscle movement. parent segment. 5). overlapping. [17]. to maintain net joint moments.

Multiple movement patterns are Jun. 21. 18087–18097. inhibitory regulation of the neck muscles.. motor out. sensory analysis. vol. Usually. hitting and kicking. 2012. AL and for their help in recruiting participants for this study. “scissor” indicates the effect of proactive-selective-inhibition targeted Specifically. automated sensory-association- regulate global motor function. 381–386. “Selective inhibition of movement. shows processes of sensory analysis. Function in Health and/or Disease allows new sensorimotor mapping. Our interpre- input and output. 10. 10 for selective inhibition within frontostriatal loops. All processes. Feedback has the potential to amplify and diminish tation of the method and results proposes these changes were the (mal)-adaptive consequences of selection. feedback countering conditions reinforced by vicious cycles ically related to the physical and psychological demands of positive feedback. 2007. Augmented feedback can assist learning voluntary.bapam. A. injury and performance limitation. particularly the shoulders. muscles. depending upon whether the feedback loop gain is positive or negative. proactive-selective inhibition of the neck muscles at the neck muscles to inhibit highly facilitated. 2013. and where positive feed- E. selection. (thicker line) pass trans-cortically. Predicts Events. we R EFERENCES outline a theoretical basis generalizing the potential of this [1] R. Shadmehr and S. M.. and R. automatic. and chinrest compression) is indeed that associated (www. Potentially. J. 4. Fig. They also thank BAPAM flexion. This [3] P. inhibitory regulation of the neck muscles has potential therapeutic value. and W. Briefly. sensory- motor associations. bypassing the slow frontal-striatal selection loops. neck. injury and performance limitation. motor inhibits unnecessary anticipatory. pp. the proposed mech- anism applies generally to all activities which engage muscles control is not apparent immediately but evolves over time. Cambridge. positive feedback within this loop has a closed loop process with selection occurring inside the feedback been shown to reinforce the development of focal dystonia. For example. and Makes Muscle action and movement operate within a feedback Decisions. loskeletal symptoms affecting mainly the upper extremities. show that direct voluntary inhibitory regulation of neck muscle can indirectly achieved by disfacilitating fast. The lation of the neck muscles has a global regulatory benefit.” Curr. J. D. the cumulative effects of biomechanical loading. 2480–2489. Different patterns also have differing cumulative cost in sympathetic arousal. and provides negative Violinists are subject to medical problems that are specif. vol. USA: MIT Press. loop dynamic system [35]. 5–9). tendons [2] J. 25. P.” J. C. Neurobiol. How the Brain Builds Representations. and back [30]. 2011. The proposed mechanism opens Regulation for Indirect Regulation of Body the loop by inhibiting the automated sensorimotor control. Neurophysiol. The ACKNOWLEDGEMENTS fact that sensorimotor control patterns are highly automated. Lutti. Proposed process model applying voluntary inhibitory neck feedback is associated with maladaptation leading to long term regulation to regulation of chronic conditions. [34]. corticostriatal circuits. pp. Smittenaar. 10) the consequences of selection are subject to feedback and will amplify or diminish through time. Nov. motor pathways and by utilizing slow. 3. linists [30]–[32]. VOL. Hence voluntary. Opin.” J. M. is relevant to high performance sports involving goal ori- and neural adaptation such as de-differentiation of receptive ented throwing. slow frontal- The new evidence of this paper is that voluntary regu- striatal loops have time for goal evaluated selection [e. The proposal is that proactive-selective. While these results apply to this study. 25]. [36]–[39]. [4] M. were not available mechanism applies to many conditions in which sensorimotor when the learning was formulated and consolidated. loop that includes the nervous system. MacDonald for their support of with chronic pain. Within this closed loop process (Fig. . They thank D. pp. loop. The results suggest a role for technology of their profession [30] with the majority reporting muscu. Coxon. The pat. NO. Frank.uk) and I. Dolan. Cortical receptive fields are known to adapt depending upon attention and the information content of the repetitive sensory input [33]. MA. movement biomechanics adapt according to their prevailing put. 97. Biological Learning and Control— result.org.g.. Selection: highly facilitated (automated) sensorimotor responses chronic pain and injury [33]. neck with custom designed equipment. positive Fig. improves global balance and reduces cost. tion and reinforcement learning [4]. “Computational models of motivated action selection in put. Stinear. This perception-action model shows sensorimotor control operating as disorder. selection. combined with the fact that musicians depend upon their The authors thank the anonymous participants for gen- learned patterns for performance and livelihood means that erously giving their time and interest for these experi- changing engrained sensorimotor control is difficult. Mussa-Ivaldi..368 IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING. and biomechanical interaction with the environment. Results (Figs. frontal striatal pathways inhibition of neck muscle can regulate vicious positive feedback leading to which allow new learning through their mechanisms of selec- pain. The Relevance of Voluntary Inhibitory Neck back reinforces the symptoms. Neurosci. crossing the neck. ments. Byblow. in training voluntary regulation of the neck muscles. Guitart-Masip. Improved global balance and reduced cost For example. “Preparing idea of a perception-action cycle is well-established. Richards for his technical support tern eliminated by neck regulation (protracted shoulder. injury and performance limitation in vio. mapping is automated and/or associated with chronic pain. Given sufficient global or selective inhibition. no. 33. vol. transient and sustained out- generation. APRIL 2017 compatible with task performance but have differing phys- iological costs including differing consequences in location and extent of biomechanical loading and fatigue. and biomechanical-environmental interaction as a closed. [25]. the proposed fields by repetitive attended behavior [33].

pp.K. Modern Musculoskeletal Physiotherapy. pp. Neurosci. 295–301.” Gait Posture. E. Schouten. 2012. D. 347–357. Lipshits. pp. [39] I. pp. G. [23] T. 463–474.” J. in Handbook Psychophysiology.: PROACTIVE SELECTIVE INHIBITION TARGETED AT THE NECK MUSCLES 369 [5] P. Lewis. analysis and visualisation of deep cervical muscle structure. 25–49. “Real-time ultrasound seg. vol. A. A. Loram. 289–306. van de Kamp. H. 519–533. USA: Elsevier. Gandevia. NY. M. Wikstrom. E. J. 464–476. vol. “Does the motor system need intermittent control?” Exerc. pp. Physiol. BMVA Symp.. H. vol. pp. Neurosci. tance: A proposal for standardization. Press. G. Sterling. 2012. Warden. 207–227. Lykken and P. Neurobiol. A. Eds. 1651–1697. 1. vol. no. 1999. Winter. Bull Rev. Loram. Blouin. 1994. 648–660. IL. Gray. Gawthrop.. vol. 1983. I. P. no. Forbes. Ishac. A. pp. C. Prinz. “The proprioceptive senses: Their roles in 2010.2623819. 1981. 25–117.. S. “A real-time biofeedback striatum and pallidum.” Repair. Proske and S. Eds. pp. 165–178. 16. Abrahams. Lederman. A. Cohen. Motor Control and Learning: A Behaviour vol. 218. “Optimal feedback control and the electromyographic study. Falla. [37] J. Costen. K.. [7] R. Using GPUs Jul. J. E.” [38] S. Lee. no. “The influence of neck-shoulder pain on trapez- Physiol. Lederman. 656–672. and G. Jan. [19] D. U. Grillner. L. Jun. 364–370.. doi: 10.2016. 2001. Saitoh. [30] A. Eds. Bosmans. D. “Sensory representation abnormalities that parallel Exerc. “Basal ganglia. Jun. “Ergonomic considerations in violists’ left [14] G. C. and F.. p. Problems Performing Artists. no. H.. “The theory of reinvestment. B. “Mechanisms for selection of basic motor programs—Roles for the [21] P. and K. De Schutter. 2012. initiation in patients with Parkinson’s disease. 1971. pp. “Models of basal ganglia. 94. 2. pp. 2000. 1633.” Trends Neurosci. Lestienne. Berque and H. vol./Feb. “Detections of movements imposed on dimensional reaching tasks. 7. vol. A. B. T. vol. Maxwell. vol. pp. [25] H. Int. 126.. postural control during upright stance and forward lean. Moore. 15. R. “Postural control and sensorimotor integration. 94. May 2012. 14. Jan. Jonkers. Jull. Mar. A. Gielo-Perczak. Ahlers. A. “Task. 2007. Sataloff. vol. Apr. vol. tive. vol.. vol. V. Pataky. [15] R.K. and R. G. no. Vis. Eds. 92. Kinesiol. Sport Sci.: Cambridge Univ. p. V. R. Feb. 1998. 1967.1109/TMI. pp. 2002. pp. Cacioppo. Oxford. vol. A. C. 2002.. p. learning: A review. A.. Helligren. pp.” J. L. “Anticipatory neck tion and role in head movement. vol.. 3. 4th ed. and muscle force. Med. Falla. Horak. Tassinary. 8. T... 2008. McCloskey. 2. “Directing attention to movement effects enhances musicians. A. Brain Res. G. I. P. Sport [33] D. Brandfonbrener. Hall and D.LORAM et al. C. Integr. R. G. C. pp. Rainoldi. R. [27] M. Menard. and I. [6] M. Masters and J. H. T. vol. 1991. Knowlton. [32] J. [29] L. 29. Eng. B. J. McCarthy.” in Performing Arts Medicine. J. Jun. Brandfonbrener. Yin and B. “Musculoskeletal problems of instrumental [17] G. no. Harding. A. vol. New York. Lett. [12] U. and P. “Attentional focus and motor learning: A review of 15 years. Blum and J. and the neck. Berntson. and F. A. pp. J. elbow and shoulder joints. Sep. 9. pp. 2005. Jull. S. Van Emmerik.” J. and J. N. T. 4. Electromyogr. and E.” Nature Rev. 68–75. Filion. Scholarpedia.” in Proc.” Psychon. 24–27. Gollee. and J.” Comput. 80. 2. Dawson. A. 25–29. Methods Biomech. Venables. 37. pp.. 2012. pp.” in Performing Arts Medicine. muscle activity associated with rapid arm movements. 3rd ed. Pruszynski and S. mentation. 17. Schoner. 42.” [31] P. Nisbet.” Neurosci. Gurfinkel. Brandfonbrener. 8. no. Science Medicine. 2012. “Direct measurement of skin conduc. Johnson and R. 159–181. USA: Human Kinetics. pp. 4. M. and M. Sataloff. Rev. 160–183. Wulf. focal hand dystonia in a primate model. 2014. “The role of the basal ganglia in habit “Lighten up: Specific postural instructions affect axial rigidity and step formation. Press. 2015. Neural [26] M. Problems Performing Artists.” shoulder pain. application using ultrasound imaging on the GPU.. “Quantifying individual muscle contribution to three. 341–359. and D. Lakie.” Neurorehabil. and M. A. 2006. Darby. “The electrodermal system. C. 135–141. 3.” in Grieve’s vol. [18] J. G. Identifying control variables for a functional task. pp.” Scholarpedia. 81–93. Rev. 2014. Blake et al. Nov. vol. vol. Science Medicine. [16] R. “Stiffness control of balance in quiet standing. J.. Oct.” Psychophysiology. pp. S. Oct. Champaign. pp. Cambridge. Redgrave. “The muscles and joints of the neck: Their specialisa- [9] V. [20] R. python.” Prog. 3. Cunningham. Imag. Loram. ius muscle activity among professional violin and viola players: An [13] J. Rev.” Motor Control. 1211–1221.” Biofizika. [35] I. A. J. “Spatio-temporal evalu. 2005. 579–584. elements of voluntary movements. J. pp.. [11] D. Hoppmann. 35. Neurophysiol. I. G. Jul. 335. long-latency stretch response. Emphasis. 2007. body position and movement. “On the control G.” Trends Neurosci. 2010. pp. vol. Scholz and G. P. 77–104. Gurfinkel.. Krzanowski. Harding. Jan. Schmidt and T. 2015.. “One-dimensional statistical parametric mapping in muscle and frequency dependent vestibular control of posture. C. Sport Exercise Psychol.” Front. vol. 1. Houk. Pal’tsev.. pp. vol. 2.” Med. IEEE Trans.” J.. Biomed. [28] V. 2007. Swinnen. U. Gurfinkel. 4th ed. Siegmund. S. Schell..” Med.. signaling body shape. 8. I.” Exp. subjects. 19. Dutia. vol. Belen’kii. . Prince. Wulf and W. “Etiologies of medical problems in perform- ation of neck muscle activation during postural perturbations in healthy ing artists. E.” Int. Scott. 1988. pp. Vandenberghe. 6. 5. Brain Res.” Somatosensory Motor Res. no. 878–888.-S. 1825. Patla. [34] R. 28. 2004. Principles of Multivariate Analysis: A Users’s Perspec. Merletti. 4. A. F. [22] W. Rev. vol. “Effect of head orientation on [24] D. 12. and [8] V. 2011. “The uncontrolled manifold concept: vol. A. 104–108.: Oxford Univ. “Sensory and motor specialization in some muscles of [10] A. M. finger. [36] P. vol. vol. Kwak. Psychol. Dec.” Exp.