Interaction of finger enslaving and error compensation in multiple finger force production

The Nature of Finger Enslaving: New Results and Their Implications

We present a review on the phenomenon of unintentional finger action seen when other fingers of the hand act intentionally. This phenomenon (enslaving) has been viewed as a consequence of both peripheral (e.g., connective tissue links and multifinger muscles) and neural (e.g., projections of corticospinal pathways) factors. Recent studies have shown relatively large and fast drifts in enslaving toward higher magnitudes, which are not perceived by subjects. These and other results emphasize the defining role of neural factors in enslaving. We analyze enslaving within the framework of the theory of motor control with spatial referent coordinates. This analysis suggests that unintentional finger force changes result from drifts of referent coordinates, possibly reflecting the spread of cortical excitation.

Number of Trials Necessary to Apply Analysis Within the Framework of the Uncontrolled Manifold Hypothesis at Different Levels of Hierarchical Synergy Control

The uncontrolled manifold hypothesis is a method used to quantify motor synergies, defined as a specific central nervous system organization that maintains the task-specific stability of motor actions. The UCM allows for inter-trial variance analysis between consecutive trials. However, despite the large body of literature within this framework, there is no report on the number of movement repetitions required for reliable results. Based on the hypothetical hierarchical control of motor synergies, this study aims to determine the minimum number of trials necessary to achieve a good to excellent level of reliability. Thirteen young, healthy participants performed fifteen bilateral isometric contractions of elbow flexion when visual feedback was provided. The force and electromyography data were recorded to investigate synergies at different levels of hierarchical control. The intraclass correlation coefficient was used to determine the reliability of the variance indices. Based on the obtained results, at least twelve trials are required to analyze the inter-trial variance in both force and muscle synergies within the UCM framework.

Aging differentially affects online control and offline control in finger force production

Human central nervous system (CNS) undergoes neurological changes during the aging process, leading to declines in hand and finger functions. Previous studies have shown that the CNS can independently process multi-finger force control and moment of force control. However, if both force and moment control are simultaneously imposed by motor task constraints, the CNS needs to resolve competing interests of generating negative and positive covariances between fingers, respectively, which causes "conflict of interest or COI". Here, we investigated how aging affects the CNS's abilities to solve COI through a new experimental paradigm. Both elderly and young subjects performed a constant force production task using index and middle fingers under two conditions, multi-finger pressing with no COI and with COI. We found that the elderly increased variance of a virtual finger (VF: an imagined finger producing the same mechanical effect as both fingers together) in time-to-time basis (i.e. online control), while increasing covariance between individual fingers (IF) forces in trial-to-trial basis (i.e. offline control) with COI than no COI. Aging affects the CNS's abilities to solve COI by deteriorating VF actions in online control and IF actions in offline control.

Timing and extent of finger force enslaving during a dynamic force task cannot be explained by EMG activity patterns

Finger enslaving is defined as the inability of the fingers to move or to produce force independently. Such finger enslaving has predominantly been investigated for isometric force tasks. The aim of this study was to assess whether the extent of force enslaving is dependent on relative finger movements. Ten right-handed subjects (22–30 years) flexed the index finger while counteracting constant resistance forces (4, 6 and 8 N) orthogonal to the fingertip. The other, non-instructed fingers were held in extension. EMG activities of the mm. flexor digitorum superficialis (FDS) and extensor digitorum (ED) in the regions corresponding to the index, middle and ring fingers were measured. Forces exerted by the non-instructed fingers increased substantially (by 0.2 to 1.4 N) with flexion of the index finger, increasing the enslaving effect with respect to the static, pre-movement phase. Such changes in force were found 260–370 ms after the initiation of index flexion. The estimated MCP joint angle of the index finger at which forces exerted by the non-instructed fingers started to increase varied between 4° and 6°. In contrast to the finger forces, no significant changes in EMG activity of the FDS regions corresponding to the non-instructed fingers upon index finger flexion were found. This mismatch between forces and EMG of the non-instructed fingers, as well as the delay in force development are in agreement with connective tissue linkages being slack when the positions of the fingers are similar, but pulled taut when one finger moves relative to the others. Although neural factors cannot be excluded, our results suggest that mechanical connections between muscle-tendon structures were (at least partly) responsible for the observed increase in force enslaving during index finger flexion.

Multi-Finger Interaction and Synergies in Finger Flexion and Extension Force Production

The aim of this study was to discover finger interaction indices during single-finger ramp tasks and multi-finger coordination during a steady state force production in two directions, flexion, and extension. Furthermore, the indices of anticipatory adjustment of elemental variables (i.e., finger forces) prior to a quick pulse force production were quantified. It is currently unknown whether the organization and anticipatory modulation of stability properties are affected by force directions and strengths of in multi-finger actions. We expected to observe a smaller finger independency and larger indices of multi-finger coordination during extension than during flexion due to both neural and peripheral differences between the finger flexion and extension actions. We also examined the indices of the anticipatory adjustment between different force direction conditions. The anticipatory adjustment could be a neural process, which may be affected by the properties of the muscles and by the direction of the motions. The maximal voluntary contraction (MVC) force was larger for flexion than for extension, which confirmed the fact that the strength of finger flexor muscles (e.g., flexor digitorum profundus) was larger than that of finger extensor (e.g., extensor digitorum). The analysis within the uncontrolled manifold (UCM) hypothesis was used to quantify the motor synergy of elemental variables by decomposing two sources of variances across repetitive trials, which identifies the variances in the uncontrolled manifold (V_UCM) and that are orthogonal to the UCM (V_ORT). The presence of motor synergy and its strength were quantified by the relative amount of V_UCM and V_ORT. The strength of motor synergies at the steady state was larger in the extension condition, which suggests that the stability property (i.e., multi-finger synergies) may be a direction specific quantity. However, the results for the existence of anticipatory adjustment; however, no difference between the directional conditions suggests that feed-forward synergy adjustment (changes in the stability property) may be at least independent of the magnitude of the task-specific apparent performance variables and its direction (e.g., flexion and extension forces).

Assessing postural stability via the correlation patterns of vertical ground reaction force components

Background

Many methods have been proposed to assess the stability of human postural balance by using a force plate. While most of these approaches characterize postural stability by extracting features from the trajectory of the center of pressure (COP), this work develops stability measures derived from components of the ground reaction force (GRF).

Methods

In comparison with previous GRF-based approaches that extract stability features from the GRF resultant force, this study proposes three feature sets derived from the correlation patterns among the vertical GRF (VGRF) components. The first and second feature sets quantitatively assess the strength and changing speed of the correlation patterns, respectively. The third feature set is used to quantify the stabilizing effect of the GRF coordination patterns on the COP.

Results

In addition to experimentally demonstrating the reliability of the proposed features, the efficacy of the proposed features has also been tested by using them to classify two age groups (18–24 and 65–73 years) in quiet standing. The experimental results show that the proposed features are considerably more sensitive to aging than one of the most effective conventional COP features and two recently proposed COM features.

Conclusions

By extracting information from the correlation patterns of the VGRF components, this study proposes three sets of features to assess human postural stability during quiet standing. As demonstrated by the experimental results, the proposed features are not only robust to inter-trial variability but also more accurate than the tested COP and COM features in classifying the older and younger age groups. An additional advantage of the proposed approach is that it reduces the force sensing requirement from 3D to 1D, substantially reducing the cost of the force plate measurement system.

Intra-Auditory Integration Improves Motor Performance and Synergy in an Accurate Multi-Finger Pressing Task

Humans detect changes in the air pressure and understand the surroundings through the auditory system. The sound humans perceive is composed of two distinct physical properties, frequency and intensity. However, our knowledge is limited how the brain perceives and combines these two properties simultaneously (i.e., intra-auditory integration), especially in relation to motor behaviors. Here, we investigated the effect of intra-auditory integration between the frequency and intensity components of auditory feedback on motor outputs in a constant finger-force production task. The hierarchical variability decomposition model previously developed was used to decompose motor performance into mathematically independent components each of which quantifies a distinct motor behavior such as consistency, repeatability, systematic error, within-trial synergy, or between-trial synergy. We hypothesized that feedback on two components of sound as a function of motor performance (frequency and intensity) would improve motor performance and multi-finger synergy compared to feedback on just one component (frequency or intensity). Subjects were instructed to match the reference force of 18 N with the sum of all finger forces (virtual finger or VF force) while listening to auditory feedback of their accuracy. Three experimental conditions were used: (i) condition F, where frequency changed; (ii) condition I, where intensity changed; (iii) condition FI, where both frequency and intensity changed. Motor performance was enhanced for the FI conditions as compared to either the F or I condition alone. The enhancement of motor performance was achieved mainly by the improved consistency and repeatability. However, the systematic error remained unchanged across conditions. Within- and between-trial synergies were also improved for the FI condition as compared to either the F or I condition alone. However, variability of individual finger forces for the FI condition was not significantly decreased as compared to I condition alone. This result indicates an improvement in motor performance is consistent with Bayesian estimation, and changes in multi-finger interaction mostly result in the enhanced motor performance. These findings provide evidence that the central nervous system can take advantage of the intra-auditory integration in a statistically optimal (Bayesian) fashion to enhance motor performance by improving multi-finger synergy.

The role of tactile sensation in online and offline hierarchical control of multi-finger force synergy

The hand, one of the most versatile but mechanically redundant parts of the human body, must overcome imperfect motor commands and inherent noise in both the sensory and motor systems in order to produce desired motor actions. For example, it is nearly impossible to produce a perfectly consistent note during a single violin stroke or to produce the exact same note over multiple strokes, which we denote online and offline control, respectively. To overcome these challenges, the central nervous system synergistically integrates multiple sensory modalities and coordinates multiple motor effectors. Among these sensory modalities, tactile sensation plays an important role in manual motor tasks by providing hand-object contact information. The purpose of this study was to investigate the role of tactile feedback in individual finger actions and multi-finger interactions during constant force production tasks. We developed analytical techniques for the linear decomposition of the overall variance in the motor system in both online and offline control. We removed tactile feedback from the fingers and demonstrated that tactile sensors played a critical role in the online control of synergistic interactions between fingers. In contrast, the same sensors did not contribute to offline control. We also demonstrated that when tactile feedback was removed from the fingers, the combined motor output of individual fingers did not change while individual finger behaviors did. This finding supports the idea of hierarchical control where individual fingers at the lower level work together to stabilize the performance of combined motor output at the higher level.

Changes in sensory function and force production in adults with type II diabetes

INTRODUCTION

The purpose of this study was to evaluate the relationship among sensory function, disease severity, and upper extremity force production in adults with type II diabetes (T2D) as compared with healthy age- and gender-matched controls.

METHODS

Ten adults with T2D and 10 healthy age- and gender-matched control subjects underwent a battery of sensory and motor function evaluations. Data on disease severity and duration were also collected.

RESULTS

The T2D group exhibited sensory deficits and altered force production as compared with healthy controls. Sensory function correlated with disease severity, as did signal predictability of kinetic output during submaximal force production tasks. Maximal force production tasks were associated with altered output in T2D, but these data did not correlate with disease severity or sensory dysfunction.

CONCLUSIONS

Some, not all, motor performance deficits in T2D are associated with sensory dysfunction. Mechanisms responsible for these changes in adult-onset T2D are described.

Movement error rate for evaluation of machine learning methods for sEMG-based hand movement classification

There has been increasing interest in applying learning algorithms to improve the dexterity of myoelectric prostheses. In this work, we present a large-scale benchmark evaluation on the second iteration of the publicly released NinaPro database, which contains surface electromyography data for 6 DOF force activations as well as for 40 discrete hand movements. The evaluation involves a modern kernel method and compares performance of three feature representations and three kernel functions. Both the force regression and movement classification problems can be learned successfully when using a nonlinear kernel function, while the exp- χ(2) kernel outperforms the more popular radial basis function kernel in all cases. Furthermore, combining surface electromyography and accelerometry in a multimodal classifier results in significant increases in accuracy as compared to when either modality is used individually. Since window-based classification accuracy should not be considered in isolation to estimate prosthetic controllability, we also provide results in terms of classification mistakes and prediction delay. To this extent, we propose the movement error rate as an alternative to the standard window-based accuracy. This error rate is insensitive to prediction delays and it allows us therefore to quantify mistakes and delays as independent performance characteristics. This type of analysis confirms that the inclusion of accelerometry is superior, as it results in fewer mistakes while at the same time reducing prediction delay.

The organization of digit contact timing during grasping

While the process of hand preshaping during grasping has been studied for over a decade, there is relatively little information regarding the organization of digit contact timing (DCT). This dearth of information may be due to the assumption that DCT while grasping exhibits few regularities or to the difficulty in obtaining information through traditional movement recording techniques. In this study, we employed a novel technique to determine the time of digit contacts with the target object at a high precision rate in normal healthy participants. Our results indicate that, under our task conditions, subjects tend to employ a radial to ulnar pattern of DCT which may be modulated by the shape of the target object. Moreover, a number of parameters, such as the total contact time, the frequency of first contacts by the thumb and index fingers and the number of simultaneous contacts, are affected by the relative complexity of the target object. Our data support the notion that a great deal of information about the object's physical features is obtained during the early moments of the grasp.

The force synergy of human digits in static and dynamic cylindrical grasps

This study explores the force synergy of human digits in both static and dynamic cylindrical grasping conditions. The patterns of digit force distribution, error compensation, and the relationships among digit forces are examined to quantify the synergetic patterns and coordination of multi-finger movements. This study recruited 24 healthy participants to perform cylindrical grasps using a glass simulator under normal grasping and one-finger restricted conditions. Parameters such as the grasping force, patterns of digit force distribution, and the force coefficient of variation are determined. Correlation coefficients and principal component analysis (PCA) are used to estimate the synergy strength under the dynamic grasping condition. Specific distribution patterns of digit forces are identified for various conditions. The compensation of adjacent fingers for the force in the normal direction of an absent finger agrees with the principle of error compensation. For digit forces in anti-gravity directions, the distribution patterns vary significantly by participant. The forces exerted by the thumb are closely related to those exerted by other fingers under all conditions. The index-middle and middle-ring finger pairs demonstrate a significant relationship. The PCA results show that the normal forces of digits are highly coordinated. This study reveals that normal force synergy exists under both static and dynamic cylindrical grasping conditions.

Dynamical degrees of freedom and correlations in isometric finger force production

Prior research has concluded that the correlations of isometric finger forces represent the extent to which the fingers are controlled as a single unit. If this is the case, finger force correlations should be consistent with estimates of the controlled (dynamical) degrees of freedom in finger forces. The present study examined the finger force correlations and the dynamical degrees of freedom in four isometric force tasks. The tasks were to produce a preferred level of force with the (a) Index, (b) Ring, (c) Both fingers and also to (d) Rest the fingers on the load cells. Dynamical degrees of freedom in finger forces were lowest in the Both finger force task and progressively higher in the Ring, Index and Resting finger force tasks. The finger force correlations were highest in the Resting and lowest in the Index and Ring finger tasks. The results for the dynamical degrees of freedom in finger forces were consistent with a reduction in degrees of freedom in response to the degrees of freedom problem and the task constraints. The results for the finger force correlations were inconsistent with a reduction in the dynamical degrees of freedom. These findings indicate that finger force correlations do not necessarily reflect the coupling of finger forces. The findings also highlight the value of time-domain analyses to reveal the organization of control in isometric finger forces.

Stabilization of the total force in multi-finger pressing tasks studied with the 'inverse piano' technique

When one finger changes its force, other fingers of the hand can show unintended force changes in the same direction (enslaving) and in the opposite direction (error compensation). We tested a hypothesis that externally imposed changes in finger force predominantly lead to error compensation effects in other fingers thus stabilizing the total force. A novel device, the "inverse piano", was used to impose controlled displacements to one of the fingers over different magnitudes and at different rates. Subjects (n=10) pressed with four fingers at a constant force level and then one of the fingers was unexpectedly raised. The subjects were instructed not to interfere with possible changes in the finger forces. Raising a finger caused an increase in its force and a drop in the force of the other three fingers. Overall, total force showed a small increase. Larger force drops were seen in neighbors of the raised finger (proximity effect). The results showed that multi-finger force stabilizing synergies dominate during involuntary reactions to externally imposed finger force changes. Within the referent configuration hypothesis, the data suggest that the instruction "not to interfere" leads to adjustments of the referent coordinates of all the individual fingers.

Multi-finger interaction during involuntary and voluntary single finger force changes

Two types of finger interaction are characterized by positive co-variation (enslaving) or negative co-variation (error compensation) of finger forces. Enslaving reflects mechanical and neural connections among fingers, while error compensation results from synergic control of fingers to stabilize their net output. Involuntary and voluntary force changes by a finger were used to explore these patterns. We hypothesized that synergic mechanisms will dominate during involuntary force changes, while enslaving will dominate during voluntary finger force changes. Subjects pressed with all four fingers to match a target force that was 10% of their maximum voluntary contraction (MVC). One of the fingers was unexpectedly raised 5.0 mm at a speed of 30.0 mm/s. During finger raising the subject was instructed "not to intervene voluntarily". After the finger was passively lifted and a new steady-state achieved, subjects pressed down with the lifted finger, producing a pulse of force voluntarily. The data were analyzed in terms of finger forces and finger modes (hypothetical commands to fingers reflecting their intended involvement). The target finger showed an increase in force during both phases. In the involuntary phase, the target finger force changes ranged between 10.71 ± 1.89% MVC (I-finger) and 16.60 ± 2.26% MVC (L-finger). Generally, non-target fingers displayed a force decrease with a maximum amplitude of -1.49 ± 0.43% MVC (L-finger). Thus, during the involuntary phase, error compensation was observed--non-lifted fingers showed a decrease in force (as well as in mode magnitude). During the voluntary phase, enslaving was observed--non-target fingers showed an increase in force and only minor changes in mode magnitude. The average change in force of non-target fingers ranged from 21.83 ± 4.47% MVC for R-finger (M-finger task) to 0.71 ± 1.10% MVC for L-finger (I-finger task). The average change in mode of non-target fingers was between -7.34 ± 19.27% MVC for R-finger (L-finger task) and 7.10 ± 1.38% MVC for M-finger (I-finger task). We discuss a range of factors affecting force changes, from purely mechanical effects of finger passive lifting to neural synergic adjustments of commands to individual fingers. The data fit a recently suggested scheme that merges the equilibrium-point hypothesis (control with referent configurations) with the idea of hierarchical synergic control of multi-element systems.