Anticipatory signals in kinematics and muscle activity during functional grasp and release

Beckers, Niek; Fineman, Richard; Stirling, Leia

doi:10.1109/bsn.2015.7299360

Cited by 5 publications

(10 citation statements)

References 32 publications

(31 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Prior work by Beckers et al [24] indicated that anticipatory signals exist in sEMG data for grasp and release actions performed by healthy individuals interacting with objects in their environment, which is consistent with with previous work [25] highlighting anticipatory cortico-cortical inputs to the motor cortex during grasping. However, Beckers et al [24] also found that the exact placement of the sEMG sensors affected signal output sufficiently, such that even small misalignments (on the order of a centimeter) changed the resulting sEMG features in statistically-significant ways, confirming the variability observed by Gazzoni et al [26] in sEMG signals detected from the forearm. This variability is of particular importance for the use of sEMG signals as control inputs because expert placement of sensors for a prosthetic or rehabilitative system is currently required for sEMG-based control of such devices.…”

Section: Introductionsupporting

confidence: 89%

“…This study is a secondary analysis of work first reported by Beckers et al [24], which tested 20 participants. Due to data collection issues, only 16 participants’ data could be used for analysis.…”

Section: Methodsmentioning

confidence: 99%

“…Voltages were recorded from the seven sEMG sensors, along with estimated times for the start of grasp and release actions as determined by calculations on marker positions and velocities from the motion capture dataset [24]. sEMG data were divided into time epochs: windows of time for which a set of features may be calculated.…”

Section: Methodsmentioning

confidence: 99%

“…The features extracted from the sEMG time series were integrated absolute value (IAV), waveform length (WFL), variance (VAR), slope sign change (SSC), zero crossings (ZC), Wilson amplitude with a threshold of 5 mV (WAMP), three equally-spaced frequency bins from 10–450 Hz (F1, F2, F3) and three equally-spaced amplitude bins of a size of 0.66 V centered around 0 V (H1, H2, H3) [8,24,31]. Features that were directly mathematically related to these (e.g., root mean square, mean absolute value) were not considered because they did not offer additional information for classification.…”

Section: Methodsmentioning

confidence: 99%

“…Studies have shown that humans pre-shape their hands when reaching for an object, with visual cues and mental models of the object’s weight and other physical features strongly influencing the associated muscle activity [23]. Prior work by Beckers et al [24] indicated that anticipatory signals exist in sEMG data for grasp and release actions performed by healthy individuals interacting with objects in their environment, which is consistent with with previous work [25] highlighting anticipatory cortico-cortical inputs to the motor cortex during grasping. However, Beckers et al [24] also found that the exact placement of the sEMG sensors affected signal output sufficiently, such that even small misalignments (on the order of a centimeter) changed the resulting sEMG features in statistically-significant ways, confirming the variability observed by Gazzoni et al [26] in sEMG signals detected from the forearm.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Classification of Anticipatory Signals for Grasp and Release from Surface Electromyography

Siu

Shah

Stirling

2016

Sensors

Self Cite

View full text Add to dashboard Cite

Surface electromyography (sEMG) is a technique for recording natural muscle activation signals, which can serve as control inputs for exoskeletons and prosthetic devices. Previous experiments have incorporated these signals using both classical and pattern-recognition control methods in order to actuate such devices. We used the results of an experiment incorporating grasp and release actions with object contact to develop an intent-recognition system based on Gaussian mixture models (GMM) and continuous-emission hidden Markov models (HMM) of sEMG data. We tested this system with data collected from 16 individuals using a forearm band with distributed sEMG sensors. The data contain trials with shifted band alignments to assess robustness to sensor placement. This study evaluated and found that pattern-recognition-based methods could classify transient anticipatory sEMG signals in the presence of shifted sensor placement and object contact. With the best-performing classifier, the effect of label lengths in the training data was also examined. A mean classification accuracy of 75.96% was achieved through a unigram HMM method with five mixture components. Classification accuracy on different sub-movements was found to be limited by the length of the shortest sub-movement, which means that shorter sub-movements within dynamic sequences require larger training sets to be classified correctly. This classification of user intent is a potential control mechanism for a dynamic grasping task involving user contact with external objects and noise. Further work is required to test its performance as part of an exoskeleton controller, which involves contact with actuated external surfaces.

show abstract

Section: Introductionsupporting

confidence: 89%

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Classification of Anticipatory Signals for Grasp and Release from Surface Electromyography

Siu

Shah

Stirling

2016

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

Quantification and visualization of coordination during non-cyclic upper extremity motion

Fineman

Stirling

2017

Journal of Biomechanics

View full text Add to dashboard Cite

There are many design challenges in creating at-home tele-monitoring systems that enable quantification and visualization of complex biomechanical behavior. One such challenge is robustly quantifying joint coordination in a way that is intuitive and supports clinical decision-making. This work defines a new measure of coordination called the relative coordination metric (RCM) and its accompanying normalization schemes. RCM enables quantification of coordination during non-constrained discrete motions. Here RCM is applied to a grasping task. Fifteen healthy participants performed a reach, grasp, transport, and release task with a cup and a pen. The measured joint angles were then time-normalized and the RCM time-series were calculated between the shoulder-elbow, shoulder-wrist, and elbow-wrist. RCM was normalized using four differing criteria: the selected joint degree of freedom, angular velocity, angular magnitude, and range of motion. Percent time spent in specified RCM ranges was used asa composite metric and was evaluated for each trial. RCM was found to vary based on: (1) chosen normalization scheme, (2) the stage within the task, (3) the object grasped, and (4) the trajectory of the motion. The RCM addresses some of the limitations of current measures of coordination because it is applicable to discrete motions, does not rely on cyclic repetition, and uses velocity-based measures. Future work will explore clinically relevant differences in the RCM as it is expanded to evaluate different tasks and patient populations.

show abstract

Utility of Inter-subject Transfer Learning for Wearable-Sensor-Based Joint Torque Prediction Models

Sloboda

Stegall

McKindles

et al. 2021

2021 43rd Annual International Conference of the IEEE Engineering in Medicine &Amp; Biology Society (EMBC)

Self Cite

View full text Add to dashboard Cite

Generalizability between individuals and groups is often a significant hurdle in model development for human subjects research. In the domain of wearable-sensorcontrolled exoskeleton devices, the ability to generalize models across subjects or fine-tune more general models to individual subjects is key to enabling widespread adoption of these technologies. Transfer learning techniques applied to machine learning models afford the ability to apply and investigate the viability and utility such knowledge-transfer scenarios. This paper investigates the utility of single-and multi-subject based parameter transfer on LSTM models trained for "sensor-tojoint torque" prediction tasks, with regards to task performance and computational resources required for network training. We find that parameter transfer between both single-and multi-subject models provide useful knowledge transfer, with varying results across specific "source" and "target" subject pairings. This could be leveraged to lower model training time or computational cost in compute-constrained environments or, with further study to understand causal factors of the observed variance in performance across source and target pairings, to minimize data collection and model retraining requirements to select and personalize a generic model for personalized wearable-sensor-based joint torque prediction technologies. I. INTRODUCTIONWearable physiological sensors may support decisionmaking, training, and device control across a variety of clinical, health monitoring, fitness, and wearable robotics tasks [1], [2], [3], [4]. The ability to generalize physiological sensing models between individuals is often a significant hurdle in model development for more widespread adoption in these domains. Among these use cases, wearable robotics (e.g. exoskeletons) have stricter requirements for temporal granularity and specificity of state sensing than other types of activity monitoring, given the need for real-time detection of user movement and intent [5]. For this work, we focus specifically on the requirements for wearable sensing models for the wearable robotics and other real-time joint dynamics estimation use cases.

show abstract

Anticipatory signals in kinematics and muscle activity during functional grasp and release

Cited by 5 publications

References 32 publications

Classification of Anticipatory Signals for Grasp and Release from Surface Electromyography

Classification of Anticipatory Signals for Grasp and Release from Surface Electromyography

Quantification and visualization of coordination during non-cyclic upper extremity motion

Utility of Inter-subject Transfer Learning for Wearable-Sensor-Based Joint Torque Prediction Models

Contact Info

Product

Resources

About