Spatial distribution of hand-grasp motor task activity in spinal cord functional magnetic resonance imaging.

Upper extremity motor paradigms during spinal cord functional magnetic resonance imaging (fMRI) can provide insight into the functional organization of the cord. Hand-grasping is an important daily function with clinical significance, but previous studies of similar squeezing movements have not reported consistent areas of activity and are limited by sample size and simplistic analysis methods. Here, we study spinal cord fMRI activation using a unimanual isometric hand-grasping task that is calibrated to participant maximum voluntary contraction (MVC). Two task modeling methods were considered: (1) a task regressor derived from an idealized block design (Ideal) and (2) a task regressor based on the recorded force trace normalized to individual MVC (%MVC). Across these two methods, group motor activity was highly lateralized to the hemicord ipsilateral to the side of the task. Activation spanned C5-C8 and was primarily localized to the C7 spinal cord segment. Specific differences in spatial distribution are also observed, such as an increase in C8 and dorsal cord activity when using the %MVC regressor. Furthermore, we explored the impact of data quantity and spatial smoothing on sensitivity to hand-grasp motor task activation. This analysis shows a large increase in number of active voxels associated with the number of fMRI runs, sample size, and spatial smoothing, demonstrating the impact of experimental design choices on motor activation.

Data Fusion-Based Musculoskeletal Synergies in the Grasping Hand.

The hypothesis that the central nervous system (CNS) makes use of synergies or movement primitives in achieving simple to complex movements has inspired the investigation of different types of synergies. Kinematic and muscle synergies have been extensively studied in the literature, but only a few studies have compared and combined both types of synergies during the control and coordination of the human hand. In this paper, synergies were extracted first independently (called kinematic and muscle synergies) and then combined through data fusion (called musculoskeletal synergies) from 26 activities of daily living in 22 individuals using principal component analysis (PCA) and independent component analysis (ICA). By a weighted linear combination of musculoskeletal synergies, the recorded kinematics and the recorded muscle activities were reconstructed. The performances of musculoskeletal synergies in reconstructing the movements were compared to the synergies reported previously in the literature by us and others. The results indicate that the musculoskeletal synergies performed better than the synergies extracted without fusion. We attribute this improvement in performance to the musculoskeletal synergies that were generated on the basis of the cross-information between muscle and kinematic activities. Moreover, the synergies extracted using ICA performed better than the synergies extracted using PCA. These musculoskeletal synergies can possibly improve the capabilities of the current methodologies used to control high dimensional prosthetics and exoskeletons.

Grasp Analysis in the Home Environment as a Measure of Hand Function After Cervical Spinal Cord Injury.

BACKGROUND: Following a spinal cord injury, regaining hand function is a top priority. Current hand assessments are conducted in clinics, which may not fully represent real-world hand function. Grasp strategies used in the home environment are an important consideration when examining the impact of rehabilitation interventions. OBJECTIVE: The main objective of this study is to investigate the relationship between grasp use at home and clinical scores. METHOD: We used a previously collected dataset in which 21 individuals with spinal cord injuries (SCI) recorded egocentric video while performing activities of daily living in their homes. We manually annotated 4432 hand-object interactions into power, precision, intermediate, and non-prehensile grasps. We examined the distributions of grasp types used and their relationships with clinical assessments. RESULTS: Moderate to strong correlations were obtained between reliance on power grasp and the Spinal Cord Independence Measure III (SCIM; P < .05), the upper extremity motor score (UEMS; P < .01), and the Graded Redefined Assessment of Strength Sensibility and Prehension (GRASSP) Prehension (P < .01) and Strength (P < .01). Negative correlations were observed between the proportion of non-prehensile grasping and SCIM (P < .05), UEMS (P < .05), and GRASSP Prehension (P < .01) and Strength (P < .01). CONCLUSION: The types of grasp types used in naturalistic activities at home are related to upper limb impairment after cervical SCI. This study provides the first direct demonstration of the importance of hand grasp analysis in the home environment.

Augmented feedback modes during functional grasp training with an intelligent glove and virtual reality for persons with traumatic brain injury.

Introduction: Physical therapy is crucial to rehabilitating hand function needed for activities of daily living after neurological traumas such as traumatic brain injury (TBI). Virtual reality (VR) can motivate participation in motor rehabilitation therapies. This study examines how multimodal feedback in VR to train grasp-and-place function will impact the neurological and motor responses in TBI participants (n = 7) compared to neurotypicals (n = 13). Methods: We newly incorporated VR with our existing intelligent glove system to seamlessly enhance the augmented visual and audio feedback to inform participants about grasp security. We then assessed how multimodal feedback (audio plus visual cues) impacted electroencephalography (EEG) power, grasp-and-place task performance (motion pathlength, completion time), and electromyography (EMG) measures. Results: After training with multimodal feedback, electroencephalography (EEG) alpha power significantly increased for TBI and neurotypical groups. However, only the TBI group demonstrated significantly improved performance or significant shifts in EMG activity. Discussion: These results suggest that the effectiveness of motor training with augmented sensory feedback will depend on the nature of the feedback and the presence of neurological dysfunction. Specifically, adding sensory cues may better consolidate early motor learning when neurological dysfunction is present. Computerized interfaces such as virtual reality offer a powerful platform to personalize rehabilitative training and improve functional outcomes based on neuropathology.

Evaluation of feature projection techniques in object grasp classification using electromyogram signals from different limb positions.

A myoelectric prosthesis is manipulated using electromyogram (EMG) signals from the existing muscles for performing the activities of daily living. A feature vector that is formed by concatenating data from many EMG channels may result in a high dimensional space, which may cause prolonged computation time, redundancy, and irrelevant information. We evaluated feature projection techniques, namely principal component analysis (PCA), linear discriminant analysis (LDA), t-Distributed Stochastic Neighbor Embedding (t-SNE), and spectral regression extreme learning machine (SRELM), applied to object grasp classification. These represent feature projections that are combinations of either linear or nonlinear, and supervised or unsupervised types. All pairs of the four types of feature projection with seven types of classifiers were evaluated, with data from six EMG channels and an IMU sensors for nine upper limb positions in the transverse plane. The results showed that SRELM outperformed LDA with supervised feature projections, and t-SNE was superior to PCA with unsupervised feature projections. The classification errors from SRELM and t-SNE paired with the seven classifiers were from 1.50% to 2.65% and from 1.27% to 17.15%, respectively. A one-way ANOVA test revealed no statistically significant difference by classifier type when using the SRELM projection, which is a nonlinear supervised feature projection (p = 0.334). On the other hand, we have to carefully select an appropriate classifier for use with t-SNE, which is a nonlinear unsupervised feature projection. We achieved the lowest classification error 1.27% using t-SNE paired with a k-nearest neighbors classifier. For SRELM, the lowest 1.50% classification error was obtained when paired with a neural network classifier.

Biorealistic Control of Hand Prosthesis Augments Functional Performance of Individuals With Amputation.

The human hand has compliant properties arising from muscle biomechanics and neural reflexes, which are absent in conventional prosthetic hands. We recently proved the feasibility to restore neuromuscular reflex control (NRC) to prosthetic hands using real-time computing neuromorphic chips. Here we show that restored NRC augments the ability of individuals with forearm amputation to complete grasping tasks, including standard Box and Blocks Test (BBT), Golf Balls Test (GBT), and Potato Chips Test (PCT). The latter two were more challenging, but novel to prosthesis tests. Performance of a biorealistic controller (BC) with restored NRC was compared to that of a proportional linear feedback (PLF) controller. Eleven individuals with forearm amputation were divided into two groups: one with experience of myocontrol of a prosthetic hand and another without any. Controller performances were evaluated by success rate, failure (drop/break) rate in each grasping task. In controller property tests, biorealistic control achieved a better compliant property with a 23.2% wider range of stiffness adjustment than that of PLF control. In functional grasping tests, participants could control prosthetic hands more rapidly and steadily with neuromuscular reflex. For participants with myocontrol experience, biorealistic control yielded 20.4, 39.4, and 195.2% improvements in BBT, GBT, and PCT, respectively, compared to PLF control. Interestingly, greater improvements were achieved by participants without any myocontrol experience for BBT, GBT, and PCT at 27.4, 48.9, and 344.3%, respectively. The functional gain of biorealistic control over conventional control was more dramatic in more difficult grasp tasks of GBT and PCT, demonstrating the advantage of NRC. Results support the hypothesis that restoring neuromuscular reflex in hand prosthesis can improve neural motor compatibility to human sensorimotor system, hence enabling individuals with amputation to perform delicate grasps that are not tested with conventional prosthetic hands.

Power Modulations of ECoG Alpha/Beta and Gamma Bands Correlate With Time-Derivative of Force During Hand Grasp.

It is well-known that motor cortical oscillatory components are modulated in their amplitude during voluntary and imagined movements. These patterns have been used to develop brain-machine interfaces (BMI) which focused mostly on movement kinematics. In contrast, there have been only a few studies on the relation between brain oscillatory activity and the control of force, in particular, grasping force, which is of primary importance for common daily activities. In this study, we recorded intraoperative high-density electrocorticography (ECoG) from the sensorimotor cortex of four patients while they executed a voluntary isometric hand grasp following verbal instruction. The grasp was held for 2 to 3 s before being instructed to relax. We studied the power modulations of neural oscillations during the whole time-course of the grasp (onset, hold, and offset phases). Phasic event-related desynchronization (ERD) in the low-frequency band (LFB) from 8 to 32 Hz and event-related synchronization (ERS) in the high-frequency band (HFB) from 60 to 200 Hz were observed at grasp onset and offset. However, during the grasp holding period, the magnitude of LFB-ERD and HFB-ERS decreased near or at the baseline level. Overall, LFB-ERD and HFB-ERS show phasic characteristics related to the changes of grasp force (onset/offset) in all four patients. More precisely, the fluctuations of HFB-ERS primarily, and of LFB-ERD to a lesser extent, correlated with the time-course of the first time-derivative of force (yank), rather than with force itself. To the best of our knowledge, this is the first study that establishes such a correlation. These results have fundamental implications for the decoding of grasp in brain oscillatory activity-based neuroprosthetics.

Control Modification of Grasp Force Covaries Agency and Performance on Rigid and Compliant Surfaces.

This study investigated how modifications in the display of a computer trace under user control of grasp forces can co-modulate agency (perception of control) and performance of grasp on rigid and compliant surfaces. We observed positive correlation (p < 0.01) between implicit agency, measured from time-interval estimation for intentional binding, and grasp performance, measured by force-tracking error, across varying control modes for each surface type. The implications of this work are design directives for cognition-centered device interfaces for rehabilitation of grasp after neurotraumas such as spinal cord and brain injuries while considering if grasp interaction is rigid or compliant. These device interfaces should increase user integration to virtual reality training and powered assistive devices such as exoskeletons and prostheses. The modifications in control modes for this study included changes in force magnitude, addition of mild noise, and a measure of automation. Significant differences (p < 0.001) were observed for each surface type across control modes with metrics for implicit agency, performance, and grasp control efficiency. Explicit agency, measured from user survey responses, did not exhibit significant variations in this study, suggesting implicit measures of agency are needed for identifying co-modulation with grasp performance. Grasp on the compliant surface resulted in greater dependence of performance on agency and increases in agency and performance with the addition of mild noise. Noise in conjunction with perceived freedom at a flexible surface may have amplified visual feedback responses. Introducing automation in control decreased agency and performance for both surfaces, suggesting the value in continuous user control of grasp. In conclusion, agency and performance of grasp can be co-modulated across varying modes of control, especially for compliant grasp actions. Future studies should consider reliable measures of implicit agency, including physiological recordings, to automatically adapt rehabilitation interfaces for better cognitive engagement and to accelerate functional outcomes.

Human neuroimaging reveals the subcomponents of grasping, reaching and pointing actions.

Although the neural underpinnings of visually guided grasping and reaching have been well delineated within lateral and medial fronto-parietal networks (respectively), the contributions of subcomponents of visuomotor actions have not been explored in detail. Using careful subtraction logic, here we investigated which aspects of grasping, reaching, and pointing movements drive activation across key areas within visuomotor networks implicated in hand actions. For grasping tasks, we find activation differences based on the precision required (fine > coarse grip: anterior intraparietal sulcus, aIPS), the requirement to lift the object (grip + lift > grip: aIPS; dorsal premotor cortex, PMd; and supplementary motor area, SMA), and the number of digits employed (3-/5- vs. 2-digit grasps: ventral premotor cortex, PMv; motor cortex, M1, and somatosensory cortex, S1). For reaching/pointing tasks, we find activation differences based on whether the task required arm transport ((reach-to-point with index finger and reach-to-touch with knuckles) vs. point-without-reach; anterior superior parietal lobule, aSPL) and whether it required pointing to the object centre ((point-without-reach and reach-to-point) vs. reach-to-touch: anterior superior parieto-occipital cortex, aSPOC). For point-without-reach, in which the index finger is oriented towards the object centre but from a distance (point-without-reach > (reach-to-point and reach-to-touch)), we find activation differences that may be related to the communicative nature of the task (temporo-parietal junction, TPJ) and the need to precisely locate the target (lateral occipito-temporal cortex, LOTC). The present findings elucidate the different subcomponents of hand actions and the roles of specific brain regions in their computation.

Detecting and classifying three different hand movement types through electroencephalography recordings for neurorehabilitation.

Brain-computer interfaces can be used for motor substitution and recovery; therefore, detection and classification of movement intention are crucial for optimal control. In this study, palmar, lateral and pinch grasps were differentiated from the idle state and classified from single-trial EEG using only information prior to the movement onset. Fourteen healthy subjects performed the three grasps 100 times, while EEG was recorded from 25 electrodes. Temporal and spectral features were extracted from each electrode, and feature reduction was performed using sequential forward selection (SFS) and principal component analysis (PCA). The detection problem was investigated as the ability to discriminate between movement preparation and the idle state. Furthermore, all task pairs and the three movements together were classified. The best detection performance across movements (79 ± 8 %) was obtained by combining temporal and spectral features. The best movement-movement discrimination was obtained using spectral features: 76 ± 9 % (2-class) and 63 ± 10 % (3-class). For movement detection and discrimination, the performance was similar across grasp types and task pairs; SFS outperformed PCA. The results show it is feasible to detect different grasps and classify the distinct movements using only information prior to the movement onset, which may enable brain-computer interface-based neurorehabilitation of upper limb function through Hebbian learning mechanisms.

Estimación en la Intención de Agarres: Cilíndrico, Esférico y Gancho Utilizando Redes Neuronales Profundas / Estimation of Hand-Grip Intention: Cylindrical, Spherical and Hook Using Deep Neural Networks

Resumen Las amputaciones de extremidades superiores pueden producir diversos grados de incapacidad en la persona afectada, esto es exacerbado aún más, si se presenta durante un periodo de su vida laboral activa, por esta razón es de importancia social el estudio de las prótesis y algoritmos que ayuden a un mejor control de estas por parte del usuario. En esta investigación, se propone una arquitectura basada en redes neuronales recurrentes del tipo Long Short-Term Memory y redes convolucionales para la clasificación de señales electromiográficas, con aplicaciones para control de prótesis de mano. La red propuesta clasifica tres tipos de agarres realizados con la mano: cilíndrico, esférico y de gancho. El modelo propuesto al ser evaluado mostró una eficiencia (accuracy) del 89 %, en contraste con una red neuronal artificial basada en capas completamente conectadas que solo obtuvo una eficiencia del 80% en la predicción de los agarres. El presente trabajo se limita solamente a evaluar la red ante una entrada de electromiograma y no se implementó un sistema de control para la prótesis de la mano. Así, una arquitectura de redes convolucionales para el control de prótesis de mano que pueden ser entrenadas con las señales del sujeto.

Abstract Upper extremities amputations can produce different disability degrees in the amputated person, this is acerbated even more, when it happens during active working life. So, for this reason, it is of social importance the study of prostheses and algorithms that help a better control of these by the user. In this research, we propose an architecture based on recurrent neural networks, called Long Short-Term Memory, and convolutional neural networks for classification of electromyographic signals, with applications for hand prosthesis control. The proposed network classifies three types of movements made by the hand: cylindrical, spherical and hook grips. The proposed model showed an efficiency (accuracy) of 89%, in contrast to an artificial neural network based on completely connected layers that only obtained an efficiency of 80% in the prediction of the hand movements. The present work is limited to evaluate the network with an electromyogram input, the control system for hand prosthesis was not implemented. Thus, an architecture of convolutional networks for the control of hand prostheses that can be trained with the signals of the subject.