[ARTICLE] Effects of a Brain-Computer Interface With Virtual Reality (VR) Neurofeedback: A Pilot Study in Chronic Stroke Patients – Full Text

Rehabilitation for stroke patients with severe motor impairments (e.g., inability to perform wrist or finger extension on the affected side) is burdensome and difficult because most current rehabilitation options require some volitional movement to retrain the affected side. However, although these patients participate in therapy requiring volitional movement, previous research has shown that they may receive modest benefits from action observation, virtual reality (VR), and brain-computer interfaces (BCIs). These approaches have shown some success in strengthening key motor pathways thought to support motor recovery after stroke, in the absence of volitional movement. The purpose of this study was to combine the principles of VR and BCI in a platform called REINVENT and assess its effects on four chronic stroke patients across different levels of motor impairment. REINVENT acquires post-stroke EEG signals that indicate an attempt to move and drives the movement of a virtual avatar arm, allowing patient-driven action observation neurofeedback in VR. In addition, synchronous electromyography (EMG) data were also captured to monitor overt muscle activity. Here we tested four chronic stroke survivors and show that this EEG-based BCI can be safely used over repeated sessions by stroke survivors across a wide range of motor disabilities. Finally, individual results suggest that patients with more severe motor impairments may benefit the most from EEG-based neurofeedback, while patients with more mild impairments may benefit more from EMG-based feedback, harnessing existing sensorimotor pathways. We note that although this work is promising, due to the small sample size, these results are preliminary. Future research is needed to confirm these findings in a larger and more diverse population.

Introduction

Stroke is a leading cause of adult long-term disability worldwide (Mozaffarian et al., 2015), and an increasing number of stroke survivors suffer from severe cognitive and motor impairments each year. This results in a loss of independence in their daily life, such as decreased ability to perform self-care tasks and decreased participation in social activities (Miller et al., 2010). Rehabilitation following stroke focuses on maximizing restoration of lost motor and cognitive functions and on relearning skills to better perform activities of daily living (ADLs). There is increasing evidence that the brain remains plastic at later stages after stroke, suggesting additional recovery remains possible (Page et al., 2004Butler and Page, 2006). To maximize brain plasticity, several rehabilitation strategies have been exploited, including the use of intensive rehabilitation (Wittenberg et al., 2016), repetitive motor training (Thomas et al., 2017), mirror therapy (Pérez-Cruzado et al., 2017), motor-imagery (Kho et al., 2014), and action observation (Celnik et al., 2008), amongst others.

Recently, growing evidence of the positive impact of virtual reality (VR) techniques on recovery following stroke has accumulated (Bermúdez i Badia et al., 2016). When combined with conventional therapy, VR is able to effectively incorporate rehabilitation strategies such as intensity, frequency, and duration of therapy in a novel and low-cost approach in the stroke population (Laver et al., 2017). However, patients with low levels of motor control cannot benefit from current VR tools due to their low volitional motor control, range of motion, pain, and fatigue. Rehabilitation for these individuals is challenging because most current training options require some volitional movement to train the affected side, however, research has shown that individuals with severe stroke may receive modest benefits from action observation and brain-computer interfaces (BCIs) (Silvoni et al., 2011).

Merging BCIs with VR allows for a wide range of experiences in which patients can feel immersed in various aspects of their environment. This allows patients to control their experiences in VR using only brain activity, either directly (e.g., movement in VR through explicit control) or indirectly (e.g., modulating task difficulty level based on workload as implicit control) (Vourvopoulos et al., 2016Friedman, 2017). This direct brain-to-VR communication can induce a sensorimotor contingency between the patient’s internal intentions and the environment’s responsive actions, increasing the patient’s sense of embodiment of their virtual avatar (Slater, 2009Ramos-Murguialday et al., 2013).

After a stroke resulting in severe motor impairments (e.g., inability to perform wrist or finger extension on the affected side), research shows that action observation combined with physical training enhances the effects of motor training (Celnik et al., 2008), eliciting motor-related brain activity in the lesioned hemisphere, leading to modest motor improvements (Ertelt et al., 2007Garrison et al., 2013). Moreover, action observation in a head-mounted VR increases motor activity in both healthy and the post-stroke brains (Ballester et al., 2015Vourvopoulos and Bermúdez i Badia, 2016a).

In addition, neurofeedback through BCIs has been proposed for individuals with severe stroke because BCIs do not require active motor control. Research on BCIs for rehabilitation has shown that motor-related brain signals are reinforced by rewarding feedback so they can be used to strengthen key motor pathways that are thought to support motor recovery after stroke (Wolpaw, 2012). Such feedback has previously shown modest success in motor rehabilitation for severe stroke patients (Soekadar et al., 2015).

The most common brain signal acquisition technology used with BCIs in stroke patients is non-invasive electroencephalography (EEG) (Wolpaw, 2012), which provide a cost-effective BCI platform (Vourvopoulos and Bermúdez i Badia, 2016b). In BCI paradigms for motor rehabilitation, EEG signals related to motor planning and execution are utilized. During a motor attempt, the temporal pattern of the Alpha rhythm (8–12 Hz) desynchronizes. The Alpha rhythm is also termed Rolandic mu or the sensorimotor rhythm (SMR) when it is localized over the sensorimotor cortices of the brain. Mu rhythms (8–12 Hz) are considered indirect indications of the action observation network (Kropotov, 2016) and reflect general sensorimotor activity. Mu rhythms are often detected with changes in the Beta rhythm (12–30 Hz) in the form of event-related desynchronization (ERD), in which a motor action is executed (Pfurtscheller and Lopes da Silva, 1999). These EEG rhythms, or motor-related EEG signatures, are primarily detected during task-based EEG (i.e., when the patient is actively moving or imagining movement) and used for neurofeedback.

Further, neurofeedback-induced changes in brain activity have also been linked to changes in brain activity at rest. That is, after training one’s brain activity using neurofeedback, the intrinsic, resting brain activity (i.e., EEG activity in the absence of a task) may also show changes. This resting brain activity can be used to assess more generalized brain changes, and baseline resting-state signatures may be used to predict recovery (Wu et al., 2015) or response to treatments (Zhou et al., 2018). When combined with neural injury information, resting EEG parameters can also help predict the efficacy of stroke therapy.

In this study, our goal was to combine the principles of virtual reality and BCIs to elicit optimal rehabilitation gains for stroke survivors. We hypothesized that merging BCIs with VR should induce illusions of movement and a strong feeling of embodiment within a virtual body via the action observation network, activating brain areas that overlap with those controlling actual movement, which is important for mobilizing neuroplastic changes (Dobkin, 2007). Using a VR-based BCI, those with severe stroke impairments can trigger voluntary movements of the virtual arm in a closed neurofeedback loop. This helps to increase the illusion of one’s own movements through the coordination between one’s intention and the observed first-person virtual action. Therefore, we developed a training platform called REINVENT, which uses post-stroke brain signals that indicate an attempt to move and then drives the movement of a virtual avatar arm, providing patient-driven action observation in head-mounted VR (Spicer et al., 2017). Our previous work using REINVENT with healthy individuals indeed showed that the combination of VR integrated into a BCI encouraged greater embodiment, and greater embodiment was related to greater neurofeedback performance (Anglin et al., 2019).

For this study, we recruited four chronic stroke survivors to undergo a longitudinal BCI-VR intervention using REINVENT to provide EEG-based neurofeedback with simultaneous EMG acquisition. We assessed intervention results using clinical measures, Transcranial Magnetic Stimulation (TMS) and Magnetic Resonance Imaging (MRI) and compared these measures before and after the intervention. The purpose of this study was twofold. First, we sought to determine whether REINVENT is feasible for stroke patients to use across repeated sessions and second, whether REINVENT might be able to strengthen motor-related brain signals in individuals with differing levels of motor impairment after stroke.[…]

 

Continue —>  Frontiers | Effects of a Brain-Computer Interface With Virtual Reality (VR) Neurofeedback: A Pilot Study in Chronic Stroke Patients | Frontiers in Human Neuroscience

Figure 1. System architecture of a closed neurofeedback loop. From left, (1) the evoked physiological responses are captured at the interfacing layer through the data acquisition clients, (2) sent to the processing layer where the signals are filtered and logged, and then, (3) the extracted features (e.g., EEG bands) are sent to the interaction layer where VR training occurs. Written permission to use this photo was obtained from the individual.

, , , , , , ,

  1. Leave a comment

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: