MR-compatible robotics to investigate human motor control

Robotic interfaces can dynamically interact with humans performing movements and can be used to study neuromuscular adaptation. Such devices can produce computer controlled dynamics during movement and measure the interaction force and movement trajectory. On the other hand, functional magnetic resonance imaging (fMRI) can provide insight into the functional organization and plasticity of the brain. Using a robotic interface in conjunction with fMRI could allow neuroscientists to investigate the brain mechanisms of manipulation and motor learning, give therapists a tool for adaptive and patient-specific rehabilitation therapies, and assist medical doctors in functional diagnostics of motor dysfunctions. However, the MR environment imposes severe safety and electromagnetic compatibility constraints on mechatronic components, and the accessible workspace around the subject is limited. In addition, interaction with human motion must be safe and gentle. This thesis investigates the MR compatibility and performances of mechatronic elements, and develops an fMRI-compatible robotic technology consisting of sensors and actuators as well as adequate safety, control and synchronization strategies. It thus becomes possible to design various robotic systems for interaction with human motion during fMRI. This novel technology is benchmarked through the realization of several MR-compatible robotic systems which are currently being used by neuroscience groups in Japan and Europe. These include a highly MR-compatible interface for wrist motion, a tactile finger stimulation device using both intrinsically compatible and electrically powered components, as well as a two-degrees-of-freedom interface to investigate the control of multi-joint arm movements. The MR compatibility of these devices is successfully tested using a protocol developed for the particularly sensitive functional MRI sequences. In a further step towards performing multi-joint arm movements during fMRI, movement-related artifacts and biomechanical constraints are examined, and movements suitable for motor control studies are identified. Preliminary experiments with the realized systems demonstrate the potential of such robotic interfaces: the brain activation patterns during mental simulation of wrist motion are different from the patterns obtained during actual movements in interaction with the interface, and activation patterns in all conditions agree with results from literature. The well-controlled conditions as well as the recorded position and force data during fMRI open up new possibilities in data analysis and may allow new insights into the brain mechanisms involved in human motor control.

Motor and cognitive brain circuits of brain-machine interactions

Silvia Marchesotti

Brain-machine interfaces (BMIs) allow the user to control an external device such as a robotic arm, a cursor, or an avatar in a virtual world through the real-time decoding of brain signals and without the involvement of the musculoskeletal system. Although BMIs hold great promise for providing motor-impaired patients with means of control and interaction with the external world, the neurocognitive mechanisms that are involved in the use of a BMI remain poorly understood. In addition, BMIs allow researchers to investigate and separate the brain processes underlying cognitive functions from those related to motor control and afferent sensory signals that reliably accompany movements. Thus, BMIs are powerful tools for basic and applied neuroscience research. The present work investigates different stages of a BMI-mediated action and targets the subjective, behavioral, and neural signals of BMI control based on electroencephalography (EEG) signals. More specifically, the main study of this thesis has focused on the development of a multimodal imaging platform that allows EEG-based BMI control inside the magnetic resonance imaging (MRI) scanner and during the acquisition of functional MRI (fMRI). This platform has enabled us to exploit both the high temporal resolution of EEG and the high spatial resolution of fMRI to precisely identify the brain mechanisms underlying BMI control and those reflecting the subjective feeling of being in control over a BMI-action (i.e., the sense of agency, SoA). In this study we found an extended cortico-subcortical network involved in operating a motor-imagery BMI. Overall BMI performance was associated with activity in a set of regions including contralateral premotor cortex and the posterior cingulate cortex. Finally, cortical midline regions and the basal ganglia were involved in the subjective sense of controlling the BMI. In a second study, we further investigated whether the ability to control a BMI relates to the ability to perform motor imagery. Despite decades of technical advances, effective BMI-control remains limited to a subset of users. We show that inter-subject variability in BMI proficiency is associated with differences in motor imagery accuracy as captured by subjective and behavioral measurements, pointing to a prominent role of kinesthetic rather than visual imagery. We also identified enhanced lateralized ÎŒ-band oscillations over sensorimotor cortices during motor imagery in high- versus low-aptitude BMI users. Finally, we developed a novel paradigm for joint BMI actions, allowing two users to be jointly engaged in BMI control; our preliminary data provide evidence in support of the hypothesis that joint actions yield BMI-performance improvement even in absence of a physical connection. Our findings also show that during joint BMI-control the SoA over the imagery-mediated actions is significantly enhanced, and is affected by BMI abilities. This work show the potential of applying a multimodal imaging approach to the field of BMI: exploiting our EEG-BMI-fMRI platform we were able to identify, the neural mechanisms involved in motor and cognitive aspects of imagery-mediated BMI. Furthermore, our results enable us to better understand the subjective and behavioral aspects of BMI-actions, and reveal strategies that could potentially reinforce BMI control. Our findings can ultimately be of relevance for the field of neuroprosthetics and make BMI more accessible to a broader range of users.

EPFL2016

Moving Brain-Controlled Devices Outside the Lab: Principles and Applications

Ricardo Andres Chavarriaga Lozano, Inaki Asier Iturrate Gil, Robert Leeb, Serafeim Perdikis

This chapter provides an overview of the functionality and the underlying principles of the brain-computer interfaces (BCI) developed by the Chair in Non-Invasive Brain-Machine Interface (CNBI) of the Swiss Federal Institute of Technology (EPFL), as well as exemplary applications where those have been successfully evaluated. Our laboratory mainly develops \emph{non-invasive} BCI systems based on electroencephalographic (EEG) signals and, thus, devoid of medical hazards, real-time, portable, relatively low-cost and minimally obtrusive. Our research is pushing forward \emph{asynchronous} paradigms offering a spontaneous, user-driven and largely ecological interaction. Furthermore, we stand on the \emph{machine-learning} way to BCI with emphasis on personalization, configurability and adaptability, coupled with \emph{mutual learning} training protocols, so that elaborate signal processing and pattern recognition methods are optimally combined with the user's learnable modulation of brain signals towards high and robust performances and universal usability. Additionally, \emph{cognitive mental state} monitoring is employed to shape or refine the interaction. \emph{Shared-control} approaches allow smart, context-aware robotics to complement the BCI channel for more fine-grained control and reduction of the user's mental workload. Last but not least, \emph{hybrid BCI} designs exploit additional physiological signals to augment the BCI modality and enrich the control paradigm, thus also exploiting potential residual capabilities of disabled end-users. The applicability and effectiveness of the aforementioned principles is hereby demonstrated in four exemplary applications evaluated with both able-bodied and motor-disabled end-users. These applications include a hybrid, motor imagery (MI)-based speller, a telepresence robot equipped with shared-control, cognitive mental state monitoring paradigms able to recognize and correct errors, and, finally, a car driving application where a passive BCI enabled on a smart car assists towards increased safety and improved driving experience. Remarkably, our results show that the performance of end-users with disabilities was similar to that of a group of healthy users, who were more familiar with the experiment and the environment. This demonstrates that end-users are able to successfully use BCI technology.

Springer Netherlands2015

Neurofeedback Using Real-Time fMRI

Rotem Roza Kopel

Over the last decade, technological advances in the field of functional magnetic resonance imaging (fMRI) have made it possible to obtain localized measures of brain activity in real-time. This allows for applications such as online quality control of the acquisition, intra-operative fMRI, brain-computer-interfaces, and neurofeedback (NFB). Real-time fMRI (rt-fMRI) NFB studies have shown that a variety of different regions of interest (ROI) can be regulated in healthy and clinical populations, and by this, set the ground to become a potential non-invasive modulation. Moreover, other advances in the field of fMRI data analysis include functional connectivity, aiming to investigate interactions between distant brain regions. These studies have the potential to inform recent interest in the human brain âconnectome.â Our work combines these two advances in the field of fMRI, and so we investigate the impact rt-fMRI NFB has on functional network organization, with the perspective of offering improved and more advanced neurofeedback. In our work, we focus on rt-fMRI NFB based on the auditory region as a single ROI, conducted with auditory stimuli on healthy subjects. In a series of papers, first, we report that subjects can downregulate activity of their right auditory cortex and that activity of the default mode network recovers via unique time courses in at least four of its components. Second, in the context of self-regulation, we utilized data-driven independent component analysis to reveal the auditory cortex as the main hub of functional connectivity changes, with changes also occurring in the brainstem, auditory pathway, and higher order areas involved in vision, attention, and working memory. Third, we utilized multivariate analyses and cross-validation to reveal 16 brain regions associated with self-regulation training. Subjects showed one of two distinct patterns of co-activation, suggesting that differences in connectivity may be related to differences in learning strategies. Finally, a meta-analysis based on several rt-fMRI NFB publications, had shown a large regulation network, including the anterior cingulate cortex, anterior insula, and the basal ganglia. This network could be involved in the underlying mechanism of self-regulation, and could improve future design of connectivity based rt-fMRI feedback. Our studies extend initial proof-of-concept studies indicating that subjects can self-regulate their own auditory cortex and go on to show that the impact of a single ROI regulation induces global network effects. In sum, we provide a comprehensive view of functional connectivity changes that occur throughout the brain with neurofeedback-based fMRI training. Those studies have the potential to advance the development of network-based rt-fMRI NFB for tinnitus clinical population, and also for other diseases that have network characteristics.

EPFL2017