Search

Recent studies have shown the feasibility of spinal cord stimulation (SCS) for motor rehabilitation. Currently, there is an increasing interest in developing closed-loop systems employing SCS for lower-limb recovery. These closed-loop systems are based on the use of neurophysiological signals to modulate the stimulation. It is known that electromagnetic stimulation can introduce undesirable noise to the electrophysiological recordings. However, there is little evidence about how electroencephalographic (EEG) or electromyographic (EMG) activities are corrupted when a trans-spinal magnetic stimulation is applied. This paper studies the effects of magnetic SCS in EEG and EMG activity. Furthermore, a median filter is proposed to ameliorate the effects of the artifacts, and to preserve the neural activity. Our results show that SCS can affect both EEG and EMG, and that, while the median filter works well to clean the EEG activity, it did not improve the contaminations of the EMG activity. The obtained results underline the need of cleaning EMG and EEG signals contaminated by SCS, which is essential for optimal closed-loop rehabilitation.

Recent studies have demonstrated the efficacy of brain-machine interfaces (BMI) for motor rehabilitation after stroke, especially for those patients with severe paralysis. However, a cerebro-vascular accident can affect the brain in many different manners, and lesions in diverse areas, even from significantly different volumes, can lead to similar or equal motor deficits. The location of the insult influences the way the brain activates when moving or attempting to move a paralyzed limb. Since the essence of a rehabilitative BMI is to precisely decode motor commands from the brain, it is crucial to characterize how lesion location affects the measured signals and if and how it influences BMI performance. This paper compares the performances of an electroencephalography (EEG)-based movement intention decoder in two groups of severely paralyzed chronic stroke patients: 14 with subcortical lesions and 14 with mixed (i.e., cortical and subcortical) lesions. We show that the lesion location influences the performance of the BMI when decoding the movement attempts of the paretic arm. The obtained results underline the need for further developments for a better individualization of BMI-based rehabilitative therapies for stroke patients.

Including supplementary information from the brain or other body parts in the control of brain-machine interfaces (BMIs) has been recently proposed and investigated. Such enriched interfaces are referred to as hybrid BMIs (hBMIs) and have been proven to be more robust and accurate than regular BMIs for assistive and rehabilitative applications. Electromyographic (EMG) activity is one of the most widely utilized biosignals in hBMIs, as it provides a quite direct measurement of the motion intention of the user. Whereas most of the existing non-invasive EEG-EMG-hBMIs have only been subjected to offline testings or are limited to one degree of freedom (DoF), we present an EEG-EMG-hBMI that allows the simultaneous control of 7-DoFs of the upper limb with a robotic exoskeleton. Moreover, it establishes a biologically-inspired hierarchical control flow, requiring the active participation of central and peripheral structures of the nervous system. Contingent visual and proprioceptive feedback about the user's EEG and EMG activity is provided in the form of velocity modulation during functional task training. We believe that training with this closed-loop system may facilitate functional neuroplastic processes and eventually elicit a joint brain and muscle motor rehabilitation. Its usability is validated during a real-time operation session in a healthy participant and a chronic stroke patient, showing encouraging results for its application to a clinical rehabilitation scenario.

Myoelectric control of rehabilitation devices engages active recruitment of muscles for motor task accomplishment, which has been proven to be essential in motor rehabilitation. Unfortunately, most electromyographic (EMG) activity-based controls are limited to one single degree-of-freedom (DoF), not permitting multi-joint functional tasks. On the other hand, discrete EMG-triggered approaches fail to provide continuous feedback about muscle recruitment during movement. For such purposes, myoelectric interfaces for continuous recognition of functional movements are necessary. Here we recorded EMG activity using 5 bipolar electrodes placed on the upper-arm in 8 healthy participants while they performed reaching movements in 8 different directions. A pseudo on-line system was developed to continuously predict movement intention and attempted arm direction. We evaluated two hierarchical classification approaches. Movement intention detection triggered different movement direction classifiers (4 or 8 classes) that were trained and tested over a 5-fold cross validation. We also investigated the effect of 3 different window lengths to extract EMG features on classification. We obtained classification accuracies above 70% for both hierarchical approaches. These results highlight the viability of classifying online 8 upper-arm different directions using surface EMG activity of 5 muscles and represent a first step towards an online EMG-based control for rehabilitation devices.

Brain-machine interfaces (BMI) can be used to control robotic and prosthetic devices for rehabilitation of motor disorders, such as stroke. The calibration of these BMI systems is of paramount importance in order to establish a precise contingent link between the brain activity related to movement intention and the peripheral feedback. However, electroencephalographic (EEG) activity, commonly used to build non-invasive BMIs, can be easily contaminated by artifacts of electrical or physiological origin. The way these interferences can affect the performance of movement intention decoders has not been deeply studied, especially when dealing with severely paralyzed patients, which often generate more artifacts by compensatory movements. This paper evaluates the effects of removing artifacts from the data used to train a BMI decoder on a dataset of 28 severely paralyzed stroke patients. We show that cleaning the training datasets reduces the global BMI performance for decoding attempts of movement. Further, we demonstrate that this performance drop especially affects the test trials contaminated by artifacts (i.e., trials that might not reflect cortical activity but noise), but not the clean test trials (i.e., trials representing correct cortical activity). This paper underlines the importance of cleaning the datasets used to train BMI systems to improve their efficacy for decoding movement intention and maximize their neurorehabilitative potential.

Low-frequency electroencephalographic (EEG) activity provides relevant information for decoding movement commands in healthy subjects and paralyzed patients. Brainmachine interfaces (BMI) exploiting these signals have been developed to provide closed-loop feedback and induce neuroplasticity. Several offline and online studies have already demonstrated that discriminable information related to movement can be decoded from low-frequency EEG activity. However, there is still not a well-established procedure to guarantee that this activity is optimally filtered from the background noise. This work compares different configurations of non-causal (i.e., offline) and causal (i.e., online) filters to classify movement-related cortical potentials (MRCP) with six healthy subjects during reaching movements. Our results reveal important differences in MRCP decoding accuracy dependent on the selected frequency band for both offline and online approaches. In summary, this paper underlines the importance of optimally choosing filter parameters, since their variable response has an impact on the classification of low EEG frequencies for BMI.

In the past few years, innovative upper-limb rehabilitation methods have been proposed for chronic stroke patients. These methods aim at functional motor rehabilitation using Brain-machine interfaces to constitute an alternate pathway from the brain to the muscles. Even in patients with absence of residual finger movements, recovery could be achieved. The extent to which these interventions are affected by individual lesion topology is yet to be understood. In this study EEG was measured in 30 chronic stroke patients during movement attempts of the paretic arm. We show that the magnitude of the event-related desynchronization was smaller in patients presenting lesions with involvement of the motor cortex. This could have important implications on the design of new rehabilitation schemes for these patients, which might benefit from carefully tailored interventions.

This paper presents a silent-speech interface based on electromyographic (EMG) signals recorded in the facial muscles. The distinctive feature of this system is that it is based on the recognition of syllables instead of phonemes or words, which is a compromise between both approaches with advantages as (a) clear delimitation and identification inside a word, and (b) reduced set of classification groups. This system transforms the EMG signals into robust-in-time feature vectors and uses them to train a boosting classifier. Experimental results demonstrated the effectiveness of our approach in three subjects, providing a mean classification rate of almost 70% (among 30 syllables).

Feedback error-related potentials are a promising brain process in the field of rehabilitation since they are related to human learning. Due to the fact that many therapeutic strategies rely on the presentation of feedback stimuli, potentials generated by these stimuli could be used to ameliorate the patient's progress. In this paper we propose a method that can identify, in real-time, feedback evoked potentials in a time-estimation task. We have tested our system with five participants in two different days with a separation of three weeks between them, achieving a mean single-trial detection performance of 71.62% for real-time recognition, and 78.08% in offline classification. Additionally, an analysis of the stability of the signal between the two days is performed, suggesting that the feedback responses are stable enough to be used without the needing of training again the user.