Scholarly article on topic 'Brain–machine interfaces in neurorehabilitation of stroke'

Brain–machine interfaces in neurorehabilitation of stroke Academic research paper on "Clinical medicine"

CC BY
0
0
Share paper
Academic journal
Neurobiology of Disease
OECD Field of science
Keywords
{"Brain–machine interface (BMI)" / Neurorehabilitation / Stroke / Robotics / "Assistive technology" / "Brain stimulation"}

Abstract of research paper on Clinical medicine, author of scientific article — Surjo R. Soekadar, Niels Birbaumer, Marc W. Slutzky, Leonardo G. Cohen

Abstract Stroke is among the leading causes of long-term disabilities leaving an increasing number of people with cognitive, affective and motor impairments depending on assistance in their daily life. While function after stroke can significantly improve in the first weeks and months, further recovery is often slow or non-existent in the more severe cases encompassing 30–50% of all stroke victims. The neurobiological mechanisms underlying recovery in those patients are incompletely understood. However, recent studies demonstrated the brain's remarkable capacity for functional and structural plasticity and recovery even in severe chronic stroke. As all established rehabilitation strategies require some remaining motor function, there is currently no standardized and accepted treatment for patients with complete chronic muscle paralysis. The development of brain–machine interfaces (BMIs) that translate brain activity into control signals of computers or external devices provides two new strategies to overcome stroke-related motor paralysis. First, BMIs can establish continuous high-dimensional brain-control of robotic devices or functional electric stimulation (FES) to assist in daily life activities (assistive BMI). Second, BMIs could facilitate neuroplasticity, thus enhancing motor learning and motor recovery (rehabilitative BMI). Advances in sensor technology, development of non-invasive and implantable wireless BMI-systems and their combination with brain stimulation, along with evidence for BMI systems' clinical efficacy suggest that BMI-related strategies will play an increasing role in neurorehabilitation of stroke.

Academic research paper on topic "Brain–machine interfaces in neurorehabilitation of stroke"

ARTICLE IN PRESS

YNBDI-03385; No. of pages: 8; 4C:

Neurobiology of Disease xxx (2014) xxx-xxx

Contents lists available at ScienceDirect

Neurobiology of Disease

journal homepage: www.elsevier.com/locate/ynbdi

Brain-machine interfaces in neurorehabilitation of stroke

Surjo R. Soekadar a,b'*, Niels Birbaumer b,c, Marc W. Slutzkyd, Leonardo G. Cohen e

a Applied Neurotechnology Lab, Department of Psychiatry and Psychotherapy, University Hospital Tübingen, Tübingen, Germany b Institute ofMedical Psychology and Behavioral Neurobiology, University ofTübingen, Tübingen, Germany c Ospedale San Camillo, IRCCS, Venice, Italy d Northwestern University, Feinberg School of Medicine, Chicago, USA e Human Cortical Physiology and Neurorehabilitation Section, NINDS, NIH, Bethesda, MD, USA

ARTICLE INFO

ABSTRACT

Article history: Received 26 May 2014 Revised 29 October 2014 Accepted 26 November 2014 Available online xxxx

Keywords:

Brain-machine interface (BMI)

Neurorehabilitation

Stroke

Robotics

Assistive technology Brain stimulation

Stroke is among the leading causes of long-term disabilities leaving an increasing number of people with cognitive, affective and motor impairments depending on assistance in their daily life. While function after stroke can significantly improve in the first weeks and months, further recovery is often slow or non-existent in the more severe cases encompassing 30-50% of all stroke victims. The neurobiological mechanisms underlying recovery in those patients are incompletely understood. However, recent studies demonstrated the brain's remarkable capacity for functional and structural plasticity and recovery even in severe chronic stroke. As all established rehabilitation strategies require some remaining motor function, there is currently no standardized and accepted treatment for patients with complete chronic muscle paralysis. The development of brain-machine interfaces (BMIs) that translate brain activity into control signals of computers or external devices provides two new strategies to overcome stroke-related motor paralysis. First, BMIs can establish continuous high-dimensional brain-control of robotic devices or functional electric stimulation (FES) to assist in daily life activities (assistive BMI). Second, BMIs could facilitate neuroplasticity, thus enhancing motor learning and motor recovery (rehabilitative BMI). Advances in sensor technology, development of non-invasive and implantable wireless BMI-systems and their combination with brain stimulation, along with evidence for BMI systems' clinical efficacy suggest that BMI-related strategies will play an increasing role in neurorehabilitation of stroke.

© 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license

(http://creativecommons.org/licenses/by/3.0/).

Introduction

Stroke is the leading cause of disability worldwide (Lopez et al., 2006). The global burden of stroke reflecting the total number of stroke survivors and loss of disability-adjusted life years (DALY) is increasing, with most of the burden in low- and middle-income countries (Feigin et al., 2014). At the same time, the number of people that depend on assistance in their daily life has drastically increased and will further accumulate in the coming decades due to demographic factors (Birbeck et al., 2014). Besides disturbances in the cognitive and affective domains, loss of motor function represents the heaviest burden of disease after stroke. While acute loss of motor function can significantly improve in the first months after stroke, further recovery is often slow or non-existent (Langhorne et al., 2011). The neurobiological reasons

* Corresponding author at: Applied Neurotechnology Lab, Department of Psychiatry and Psychotherapy, University Hospital Tübingen, Tübingen, Germany.

E-mail addresses: surjo.soekadar@uni-tuebingen.de (S.R Soekadar), niels.birbaumer@uni-tuebingen.de (N. Birbaumer), mslutzky@northwestern.edu (M.W. Slutzky), cohenl@ninds.nih.gov (L.G. Cohen).

Available online on ScienceDirect (www.sciencedirect.com).

for this slowdown are incompletely understood and subject of intensive investigation (Burke and Cramer, 2013; Buma et al., 2013).

Currently, three main mechanisms are thought to contribute to stroke recovery. The first mechanism relates to the reduction in edema and a process termed diaschisis, i.e. a sudden loss in function with reduced blood flow and metabolism of brain areas connected to an irreversibly damaged injury core that may in part reverse in the early phase after stroke (Feeney and Baron, 1986). The second mechanism relates to functional recovery due to compensation (Lang et al., 2006; Cirstea and Levin, 2000) based on improved use and refinement of remaining motor functions. The third postulated mechanism assumes "real" recovery, i.e. restoration of lost brain functions due to homeostatic and learning-dependent reorganization of the brain (Nudo and Milliken, 1996). To different degrees, the latter two mechanisms involve or may lead to changes in neurotransmitter concentrations, neuro- and synaptogenesis, dendritic branching and axonal sprouting (Buma et al., 2013). This makes interpretation of neurophysiological or neuroimag-ing measures that strive for a strict differentiation of these different mechanisms and their dynamic interaction across stroke recovery rather challenging.

The conventional clinical wisdom used to be that by six months to a year after stroke, the potential for recovery has substantially diminished

http://dx.doi.org/10.1016/j.nbd.2014.11.025

0969-9961/© 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/3.0/).

ARTICLE IN PRESS

2 S.R. Soekadar et al. / Neurobiology of Disease xxx (2014) xxx-xxx

(an opinion reflected in many health insurance policies). This view was challenged by a meta-analysis providing an evidence base for stroke rehabilitation even in the "chronic" stage of stroke (Teasell et al., 2014). Clinical studies showed that e.g. constrained induced movement therapy (CIMT) can be effective even in the rehabilitation of chronic stroke (Sirtori et al., 2009). In CIMT, the healthy arm is constrained, which forces the patient to use the non-used paralyzed limb (Taub et al.,

2002). The success of such a strategy indicates that the degree of neural plasticity and motor learning does not entirely depend on the time after stroke, but depends in large part on learning and environmental conditions.

However, many stroke survivors (about 30-50%) do not qualify for CIMT (Wolf et al., 1989; Taub et al., 1999) as it requires remaining residual movement. For these patients, there is currently no standardized or accepted treatment strategy.

Recent advances in neurotechnology have led to the development of brain-computer or brain-machine interfaces (BCIs/BMIs) that translate electric, magnetic or metabolic brain activity into control signals of computers or machines, e.g. neuroprosthetic or robotic devices (Venkatakrishnan et al., 2014). Recent studies suggest that BMIs will become an important component of several new strategies that strive to overcome severe stroke-related motor impairments.

Brain-machine interfaces (BMIs)

Currently, there are two main strategies pursued to restore function after stroke using BMIs. The first strategy aims at bypassing nonfunctional cortico-spinal pathways to allow for continuous and permanent control of robotic devices (Collinger et al., 2013) or functional electric stimulation (FES) of paralyzed muscles (Moritz et al., 2008; Pohlmeyer et al., 2009; Ethier et al., 2012; McGie et al., in press; Pfurtscheller et al.,

2003). By substituting for lost motor functions, such assistive BMIs have demonstrated recovery of versatile motor control in daily life activities (Hochberg et al., 2006; Collinger et al., 2013). The second strategy aims at facilitation of neuroplasticity and motor learning to enhance motor recovery (rehabilitative BMIs) (Dobkin, 2007; Soekadar et al., 2011a) (Fig. 1a).

While deriving from different research traditions, both strategies probably involve the same neural mechanisms for BMI learning and control, mainly operant conditioning and feedback learning independent of the invasiveness of the approach and both involve the cortico-striatal loop (Koralek et al., 2012). In non-invasive BMIs, six types

of brain signals have been tested: 1. sensori-motor rhythms (SMR, 8-15 Hz, also termed rolandic alpha or mu-rhythm depending on the context) (McFarland et al., 1993, 2006; Pfurtscheller et al., 2006; Soekadar et al., 2011a, in press-a), 2. slow cortical potentials (SCP) (Birbaumer et al., 1999), 3. event-related potentials (ERPs) (Farwell and Donchin, 1988) and 4. steady-state visually or auditory evoked potentials (SSVEP/SSAEP) (Sakurada et al., 2013), 5. blood-oxygenation level dependent (BOLD)-contrast imaging using functional MRI (Weiskopf et al., 2003), and 6. concentration changes of oxy/deoxy hemoglobin using functional near-infrared spectroscopy (fNIRS; Sitaram etal., 2009; Miharaetal., 2013; Rea et al., 2014). Implantable BMIs, in contrast, require surgical implantation of epidural, subdural, or intracortical electrode arrays. In order to make assistive BMIs reliable in daily life environments, stable decoding of brain activity for controlling a high degree-of-freedom (DOF) output is necessary, an issue currently only achievable using invasive recordings. Implantable BMIs have successfully used local field potentials (LFPs) inside the cortex (Hwang and Andersen, 2009; Flint et al., 2013) or on the surface (Leuthardt et al., 2004; Schalk et al., 2008; Wang et al., 2013) and action potentials (spikes) (e.g., Taylor et al., 2002; Serruya et al., 2002; Carmena et al., 2003).

The first clinically relevant assistive BCI was used by patients suffering from locked-in syndrome, a condition in which patients are awake and cognitively aware of their environment, but unable to move or to speak. An EEG-based BCI translated purposeful modulation of SCP into binary selections of letters or words on a screen (Birbaumer et al., 1999, 2014). Recently, such system was successfully used in a tetraplegic patient with brainstem stroke (Sellers et al., 2014). While restoration of communication was the only relevant BMI application for a long time, the impact and relevance of both non-invasive and implantable assistive BMIs for restoration of movement were negligible. However, recent demonstrations of reliable control of a robotic arm after intracortical electrode array implantation allowing tetraplegic individuals, e.g. after brain stem stroke, to perform skillful and coordinated reaching and grasping movements (Hochberg et al., 2012; Collinger et al., 2013) generated considerable enthusiasm (see video demonstrations 1 & 2). Still, surgical implantation of hardware entails relatively low but substantial risk of infection and hemorrhage that many stroke survivors may not be willing to accept. At the same time, hybrid systems merging EEG with other biosignals, e.g. electrooculograms (EOG) and electromyograms (EMG) (Millân et al., 2010), termed brain/ neural computer interaction (BNCI) systems, have provided some remarkable examples of restored motor function, e.g., driving a wearable

a MEG Sensors . .. r"\

Biosignal recording k/

Fig. 1. a: Illustration of a brain-machine interface (BMI) system for stroke neurorehabilitation training. Bio-signals associated with attempted movements of the paralyzed hand and fingers are translated into online feedback and/or brain-state dependent transcranial electric stimulation to augment neuroplasticity facilitating motor recovery. b: Illustration of the setup used in Soekadar et al. (in press-b) to investigate Hebbian learning to control brain oscillatory activity.

ARTICLE IN PRESS

S.R. Soekadar et al. / Neurobiology of Disease xxx (2014) xxx-xxx

exoskeleton opening and closing the paralyzed hand and fingers (Soekadar et al., in press-a) to perform activities of daily living (ADLs).

These impressive demonstrations suggest that assistive BMIs will become a realistic option to improve living conditions of patients with paralysis once the associated costs and risks of these systems can be balanced with long-term benefits for the patients.

The theoretical concept of rehabilitative BMIs, also termed biofeedback or restorative BMI (Soekadar et al., 2011a), is based on the early work of Barry Sterman et al. (1969), and postulates that operant conditioning of neural activity can alter behavior. Sterman showed that operant conditioning of sensorimotor rhythms (SMRs) in patients with severe epilepsy can reduce frequency of grand-mal seizures (Sterman and Macdonald, 1978). The relevance for other neurological and psychiatric disorders of such approach was later demonstrated in controlled clinical studies, e.g. for attention deficit and hyperactivity disorder (Lubar and Shouse, 1976; Monastra et al., 2005; Strehl et al., 2006) or depression (Linden et al., 2012).

BMIs in stroke neurorehabilitation

An early case study suggested that operant conditioning of ipsilesional SMR may be beneficial after stroke (e.g., Rozelle and Budzynski, 1995). Further studies indicated that ipsilesional cortical function early after stroke predicted subsequent motor recovery (Platz et al., 2002; Calautti et al., 2010). Motivated by these results as well as previous work by Basmajian (1981), Basmajian et al. (1982), Birbaumer and Cohen (2007) developed a SMR-based BMI enabling severely affected stroke patients to control an orthotic device and thereby open or close their paralyzed hand. This system provided immediate sensory feedback contingent upon their ipsilesional brain activity (Buch et al., 2008). They hypothesized that by re-establishing contingency between ipsilesional cortical activity related to motor planning of, or attempted execution of, finger movements and proprioceptive (haptic) feedback, such a BMI might strengthen the ipsilesional sensorimotor loop and foster neuroplasticity that facilitates motor recovery (Dobkin, 2007; Birbaumer and Cohen, 2007). The hypothesized mechanism behind such plasticity involves simultaneous activation of inputs and outputs to motor cortices, thus triggering Hebbian plasticity (Fig. 2). Other groups have also tried BMI without haptic feedback using SMR as a method to monitor and train motor imagery (Prasad et al., 2010). The mechanism by which such BMIs might improve motor

Fig. 2. Schematic of hypothesized mechanism of plasticity generated using a BMI. Brain signals (field potentials) are used by the BMI to control a prosthetic device or neuromuscular stimulation, which provides somatosensory feedback to primary motor cortex (M1) via both somatosensory cortex (S1) and direct thalamic input. The simultaneous activation of presynaptic inputs to M1 with postsynaptic M1 activation causes Hebbian potentiation analogous to spike-timing dependent plasticity.

function is less clear, possibly involving attempts to return brain activity "closer to normal" (Daly and Wolpaw, 2008).

Although known for many decades, the functional role of sensorimotor cortex oscillations is still not well understood. While generated by LFP within the motor cortical areas of non-human primates (Sanes and Donoghue, 1993), SMR and beta rhythms showed no specific contingency to an actual motor output suggesting a functional relatedness with rather unspecific sensorimotor integration (Murthy and Fetz, 1996). But due to its relatedness to motor activity, accessibility by EEG and high signal-to-noise ratio, SMR seemed an ideal candidate for non-invasive BMl-training in stroke neurorehabilitation (Soekadar et al., 2011a,b).

An initial study indicated that the majority of chronic stroke patients can learn to control ipsilesional SMR (Buch et al., 2008), but a few weeks of training did not result in any significant motor function improvement or generalization of the skill into activities of daily living. However, daily BMI training coupled with goal-directed behavioral physical therapy over a longer period led to remarkable improvements of motor and cognitive capacities of a stroke survivor with severe chronic paralysis after a thalamic hemorrhage (Broetz et al., 2010) as measured by the Fugl-Meyer Assessment (FMA), Wolf Motor Function Test (WMFT) and Goal Attainment Score (GAS). While the participant was unable to use his hand or arm for any relevant activities of daily living and entirely depended on assistance for personal hygiene and dressing, all parameters of motor function improved over the course of the training. Furthermore, the patient became independent of any walking aid or assistance for personal hygiene, and concentration and attentiveness improved significantly. A longitudinal fMRI study indicated that clinical improvements were associated with an increased activation of the ipsilesional hemisphere (Caria et al., 2011). Another study that applied combined BMI and functional electric stimulation (FES) of paralyzed finger muscles in a chronic stroke survivor reported restored individual finger extension (measured as degrees of isolated index finger joint extension) after nine sessions (Daly et al., 2009; Wang et al., 2010). Encouraged by these findings, a larger clinical trial with 32 chronic stroke survivors without residual finger movements (12.22 ± 1.51 out of 54 points according to a combined hand and modified arm FMA, cFMA, indicating severe upper-limb motor impairment) was conducted and showed that motor improvements after 20 sessions of ipsilesional BMI training combined with goal-directed behavioral physiotherapy were superior to motor improvements in a sham BMI-group who received random BMI-feedback (Ramos-Murguialday et al., 2013). cFMA improvements in the BMI-group were associated with changes in fMRI laterality index and paretic hand EMG activity. Importantly, neurophysiological assessment indicated that motor recovery was correlated with the presence of upper-limb motor evoked potentials (MEPs) elicited from the ipsilesional hemisphere (Brasil et al., 2012), underlining the importance of the descending corticospinal tract's integrity for training-related recovery (Jung et al., 2012). Integrity of the ascending sensory pathways showed similar relevance for successful BMI control and learning (Shaikhouni et al., 2013). A more recent clinical study comparing conventional robot-assisted therapy with BMI-controlled robotic training in chronic stroke survivors found similar results (Ang et al., in press). While patients improved under both training conditions, re-assessment of motor function three months later indicated that more participants in the BMI-training group attained motor gains (as measured by FMA) than in the group that received conventional robot-assisted training. Other less controlled studies with smaller samples further corroborate this finding (Varkuti et al., 2013; Mukaino et al., 2014).

While these first results need confirmation through larger clinical trials, the reported outcomes are remarkable and underline the capacity of chronic stroke patients with severe motor deficits to regain motor function under effective learning conditions. As some of the reported studies included individuals with residual voluntary motor function, it is unclear how much residual function and motor pathway connectivity are needed for improved rehabilitation outcomes. Also, many

ARTICLE IN PRESS

4 S.R. Soekadar et al. / Neurobiology of Disease xxx (2014) xxx-xxx

commonly used instruments for clinical evaluation of motor function after stroke are not optimal for assessing motor recovery after severe forms of stroke as they cannot reliably differentiate minimal residual hand function (Broetz et al., 2014). This may in part explain the heterogeneity in clinical tests that were used to investigate efficacy of BMI in neurorehabilitation of stroke making direct comparisons between study results difficult (for overview see Table 1).

Based on the same principle as MEG/EEG-based BMI training, real-time fMRl (rt-fMRl) and fNIRS neurofeedback have also been used to increase the activity of ipsilesional motor cortical areas (Sitaram et al., 2012; Mihara et al., 2013). Allowing feedback of deeper brain structures, e.g. dopaminergic mid-brain regions, rt-fMRl may become an effective non-invasive tool to study the role of sub-cortical brain structures in the context of stroke recovery (Sulzer et al., 2013). Also, multisite rt-fMRl BMl feedback could be used to increase the connectivity between functionally associated brain regions (Ruiz et al., 2014).

While the majority of stroke patients, particularly those with subcortical lesions, were able to learn SMR-based BMI control (Buch et al., 2008, 2012), such BMI learning is often slower after stroke compared to healthy controls (Soekadar et al., 2011b). Thus, development of strategies aiming at enhancement of BMl learning may further increase applicability of BMl training protocols for stroke neurorehabilitation. A particularly promising tool in this context is the application of noninvasive brain stimulation (NIBS) (Dayan et al., 2013; Liew et al., 2014).

Combination of BMIs and brain stimulation in neurorehabilitation of stroke

While it was shown that the application of electric currents to the brain can modulate mood, cognition and behavior, only the recent development of neurophysiological and neuroimaging tools allows systematic investigation of the mechanisms underlying these effects (Bolwig, 2014). Besides invasive stimulation techniques, such as deep brain stimulation or motor cortex stimulation, non-invasive forms of brain stimulation (NIBS), including transcranial direct current stimulation (tDCS) or transcranial magnetic stimulation (TMS) are increasingly used (Liew et al., 2014) and their effects on brain physiology are being investigated (Dayan et al., 2013). For instance, it was shown that tDCS, i.e. the application of weak electric direct currents (DC) of 1-2 mA through saline soaked sponges or electrodes, can improve learning and consolidation throughout different domains (Reis et al., 2008; Marshall et al., 2004). When applied over the ipsilesional motor cortex of chronic stroke patients, reaction time and pinch force of the affected hand improved (Hummel et al., 2006). Similarly, facilitatory repetitive TMS (rTMS) applied to the ipsilesional hemisphere (Khedr et al., 2010) and inhibitory rTMS targeting the contralesional hemisphere (Takeuchi et al., 2005) or their combination (Sung et al., 2013) showed effects on motor functions in stroke, but more studies with larger sample sizes are needed (Hao et al., 2013).

Recently, it was shown that tDCS can enhance learning to control an SMR-based BMI (Soekadar et al., in press-b). In this study, healthy participants engaged in SMR-BMl control directly after receiving 20 min of anodal or cathodal tDCS over their primary motor cortex (M1) (Fig. 1b). After one week of daily training, improvement of SMR control was superior in those participants who received anodal tDCS compared to those who received cathodal or sham stimulation. One month after the end of the training, the newly acquired skill remained superior in the group that received anodal tDCS.

Several studies indicated that timing of tDCS relative to training can influence stimulation effects (Pirullietal., 2013; Stagg et al., 2011; Galea and Celnik, 2009; Volpato et al., 2013). Thus, development of new strategies allowing for simultaneous or state-dependent brain stimulation during BMl control promised to improve applicability and effectiveness of BMI training protocols in patients with brain lesions. Recently, successful combination of simultaneous tDCS and EEG-based BMl

control was demonstrated (Soekadar et al., 2014). However, this setup allows placing the stimulation electrode only as close as 1 cm near the EEG electrode used for BMI control to avoid direct contact resulting in amplifier saturation by stimulation currents entering the EEG system. This limits the possibility of applying electric currents near electrodes used for BMI control. Another strategy, however, uses neuromagnetic brain signals (MEG) that can pass through the stimulation electrode. This new strategy allows for in vivo assessment of neuromagnetic brain oscillations in brain regions immediately underneath the stimulation electrode (Soekadar et al., 2013a). Soekadar et al. (2013b) showed for the first time that a BMI utilizing SMR of the primary motor cortex (M1) could control an orthotic device while this region, the ipsilesional M1 of a chronic stroke survivor without residual movements, underwent anodal tDCS (Fig. 1a). This new strategy may lead to the refinement of existing stimulation protocols to improve their effectiveness and shed light on the relationship between brain physiology, cognition and behavior.

Current challenges and future developments

Thus far, almost all BMIs used for rehabilitation of stroke have been non-invasive. Yet having provided remarkable results, non-invasive measures limit the possibility to use brain signals from small generator volumes that oscillate at high frequencies due to the distance of the electrodes or sensors from the signal source. While low-frequency rhythms, such as SMR, showed limited correlation and contingency with intended movements (Schalk et al., 2007; Mehring et al., 2004; Stark and Abeles, 2007; Flint et al., 2012a), decoding of high gamma band and action potentials (single- or multi-unit) (Mehring et al., 2004; Stark and Abeles, 2007; Flint et al., 2012b) allowed for control of high degree of freedom prosthetic limbs (Hochberg et al., 2012; Collinger et al., 2013) or FES (Ethier et al., 2012).

Since the basis for neural plasticity in BMIs is hypothesized to be Hebbian plasticity involving simultaneous activation of pre- and postsynaptic neurons (Fig. 2), this contingency may be critical to driving functional connectivity optimally. Therefore, the use of other brain signals, such as high gamma/broadband power (70 to ~300 Hz) may prove equally or even more effective than lower frequency signals such as SMR. Recently, it was shown that high gamma signals can be used to decode highly fractionated movements, for example in biomi-meticBMls (Flint etal., 2013). Such signals are most effectively obtained using invasive recordings with intracortical, subdural or epidural electrodes (Mehring et al., 2004; Stark and Abeles, 2007; Zhuang et al., 2010; Slutzky et al., 2010, 2011; Flint et al., 2014). While intracranial electrodes require implantation, epidural or subdural electrodes could ultimately be implanted through a burr hole instead of a craniotomy reducing the perioperative risk and cost.

A substantial barrier here is the lack of fully implantable (and ideally, wireless) intracranial devices. One helpful development may be the recent approval of a recording and stimulation system for epilepsy (Heck et al., 2014), although this device only uses a small number of electrodes for recording. Once fully internalized systems are available, the risk-benefit ratio may change considerably. Similarly, as noninvasive brain recording technology advances, other means of providing comparable functional improvement without surgery may change the risk-benefit ratio. Ultimately, the decision to implant and apply such BMI system will be highly individual and dependent upon the patient's circumstances.

A recent study showed effective use of an implantable neural interface to bridge damaged neural pathways to restore function and promote recovery after brain injury in a rat model (Guggenmos et al., 2013). In this study, the primary motor cortical area of a rat was injured leading to a disruption of communication between motor and somatosensory areas. An implanted neural prosthesis translated action potentials in premotor cortex into contingent electrical stimulation in somatosensory cortex. After continuous application over 2 weeks,

Clinical studies that investigated brain-machine interfaces (BMIs) in stroke neurorehabilitation. Only those studies are listed that assessed motor function or electromyographic (EMG) activity before and after BMI training. EEG: electroencephalography, MEG: magnetoencephalography, fNIRS: functional near-infrared spectroscopy, fMRI: functional magnetic resonance imaging, EMG: electromyography, SMR: sensorimotor rhythm, ERD: event-related desynchronization, FES: functional electric stimulation, rBSl: revised brain symmetry index, LI: laterality index, FCC: functional connectivity correlate, FMA: Fugl-Meyer Assessment, cFMA: combined hand and modified armFMA, MAL AOU: Motor Activity Log Amount of Use, TUG: Timed Up and Go test, SIAS: Stroke Impairment Assessment Set, ARAT: Action Research Arm Test, HWC: Holden Walking Classification, MAS: Motor Assessment Scale for stroke, 9-HPT: 9 Hole Peg Test, WMFT: Wolf Motor Function Test, GAS: Goal Attainment Score, GS: grip strength, MRC: Medical Research Council Scale for Muscle Strength, Ml: motor imagery, ME: motor execution.

Number of stroke patients included

Stroke severity at inclusion

methodology

Feedback

Clinical

scores/neurophysiological measures

Ang et al. (in press) (upper extremity) Ono et al. (2014) (upper extremity)

Young et al. (2014) (upper extremity)

Ramos-Murguialday et al. (2013)

(upper extremity) Mihara et al. (2013) (upper extremity)

Varkuti et al. (2013) (upper extremity)

Takahashi et al. (2012) (lower extremity)

Shindo et al. (2011) (upper extremity)

Caria et al. (2011) (upper extremity) Broetz et al. (2010) (upper extremity) Ang et al. (2010) (upper extremity) Prasad et al. (2010) (upper extremity) Sun et al. (2011) (lower extremity) Ang et al. (2009) (upper extremity) Daly et al. (2009) (upper extremity)

Buch et al. (2008) (upper extremity)

n = 11 n = 15 n = 6 n = 6

n = 8 n = 6 n = 16 n = 16 n = 10 n = 10 n=6 n = 3 n=1

n = 11 n = 14 n=5

n = 20

n=6 n = 7 n=1

Moderate to severe EEG (SMR, beta), Ml

FMA: 26.4 ±14.8 (out of 66)

Severe EEG (SMR), ME

SlAS finger score: 0-1 (out of 5)

Mild to severe EEG (SMR, beta), ME

ARAT: 23.71 ± 25.68 (out of 57)

Severe EEG (SMR), ME

cFMA: 12.15 ± 8.8 (out of 54)

Severe fNIRS, Ml

FMAhand: <5.0 (out of 12)

Moderate to severe EEG (SMR, beta), Ml

FMA: 22.57 ± 15.2 (out of 66)

Severe EEG (beta), ME

SlAS foot tap score: 0-1 (out of 5)

Severe EEG (SMR), Ml

SIAS fingertest: 1a-2 (out of 5)

Severe EEG/MEG (SMR), Ml

FMA: 13.0 (out of 66)

Severe EEG/MEG (SMR), Ml

FMA: 13.0 (out of 66)

Severe EEG (SMR, beta)

FMA: 14.9 (out of 66)

Mild to severe EEG (SMR), Ml

ARAT: 22.6 ± 22.6 (out of 57)

Moderate to severe EEG (SMR, beta), Ml

HWC: 2.5 ±0.51 (out of 6)

Moderate to severe EEG (SMR, beta), Ml

FMA: 29.7 (out of 66)

Moderate to severe EEG (SMR), Ml

Volitional partial movement of mass finger

and thumb extension, no isolated finger movement

Severe MEG (SMR), Ml

Finger extension weakness rated as 0 out of 5 on the

MRC scale

Visual, proprioceptive/haptic (robot)

Visual, proprioceptive/haptic (orthosis)

Visual, FES-related muscle contraction

Visual, proprioceptive/haptic

(orthosis)

Visual

Visual, proprioceptive/haptic (robot)

Visual, FES-related muscle contraction

Visual, proprioceptive/haptic (orthosis)

Visual, proprioceptive/haptic (orthosis)

Visual, proprioceptive/haptic (orthosis)

Visual, proprioceptive/haptic

(robot)

Visual

Visual

Visual, proprioceptive/haptic (robot)

Visual, FES-related muscle contraction

Visual, proprioceptive/haptic (orthosis)

FMA, rBSI EMG activity, ERD

Stroke Impact Scale, ARAT, 9-HPT,

LI, fMRI

FMA, ARAT

FMA, FCC based on fMRI

EMG activity, maximum range of motion (ROM)

SIAS, MAL AOU, MAS FMA

FMA, WMFT, Ashworth Scale, GAS FMA

ARAT, GS, motoricity index, 9-HPT

HWC, Berg Balance Scale

Degrees of isolated finger extension MRC

ARTICLE IN PRESS

6 S.R. Soekadar et al. / Neurobiology of Disease xxx (2014) xxx-xxx

reaching and grasping functions have improved to a degree indistinguishable from pre-lesion levels. Similarly, ECoG signals could be linked to direct stimulation of the spinal cord's anterior horn (Nishimura et al.,

2013), e.g. in rehabilitation of sub-cortical stroke.

Although there is increasing evidence for the efficacy of BMl-related tools to improve rehabilitation strategies in stroke or other neurologic disorders, more and larger clinical studies are needed. ln particular, it is critical to investigate the underlying mechanisms of BMl-induced functional recovery. For example, does such functional improvement result from functional or structural brain reorganization, or is it merely due to compensatory mechanisms? Understanding the mechanisms of recovery could also lead to identification of biomarkers that could predict treatment response (e.g., Buch et al., 2012; Brasil et al., 2012). In addition, as true with most rehabilitative paradigms, the optimal dosage (frequency and intensity) of BMI training needs to be investigated. Advances in the field of telerehabilitation allowing for monitoring physiological parameters and treatment course may facilitate application of BMls in home-based rehabilitation programs. ln cases in which ipsilesional BMI training is not feasible due to the absence of brain signals to be trained and where the possibility to augment such signal by brain stimulation or other means is unavailable, training of contralesional, ipsilateral brain activity might be a possible alternative (Bundy et al., 2012; Carmena, 2013).

Another challenge is to identify and provide optimal frameworks for generalization of skills learned in the lab or hospital to daily life environments. Full implantation or hybrid systems combining features of rehabilitative and assistive BMls may facilitate generalization and stabilize regained motor function. This could reduce the necessity of physiotherapy or other means to increase generalization.

Finally, an important aspect of stroke that is crucial to recovery, but often neglected, is the prevalence of depression and sustained learned helplessness. Early after stroke, over 30% of all stroke victims exhibit symptoms of depression, anxiety, fatigue and apathy (Hackett et al.,

2014) impeding their motivation and capacity to engage in rehabilitation measures. lt was shown that these symptoms relate to cortico-striatal connectivity (Shepherd, 2013) and disruption of the cortico-striatal thalamocortical loop (Terroni et al., 2011). Besides application of antidepressants (e.g. fluoxetine) that showed to significantly improve clinical outcome when administered early after stroke (Chollet et al., 2011, 2014; Mead et al., 2013), techniques aiming at purposeful modulation of this circuit's activity, such as rt-fMRl or DBS targeting subcortical activity, or BMIs targeting cortical activity, may fill the current gap and help restore motor, cognitive and affective function in these stroke survivors.

Conclusions

Brain machine interfaces are powerful tools that can enable stroke survivors to regain movement. While larger clinical studies are needed to understand BMl-related stroke recovery mechanisms, predictors of treatment response, as well as reliability and safety of implantable systems, BMl technology is evolving towards a potentially broadly applied and important component in rehabilitation strategies for stroke survivors for whom no other treatment options exist. Combination of BMIs with invasive and non-invasive brain stimulation (NIBS) promises to provide a better understanding of mechanisms underlying brain recovery and to improve efficacy of BMIs in stroke neurorehabilitation.

Acknowledgments

This work was supported by the Intramural Research Program (IRP) of the National Institute of Neurological Disorders and Stroke (NINDS), USA; the German Federal Ministry of Education and Research (BMBF, grant number 01GQ0831, 16SV5838K to SRS and NB); the BMBF to the German Center for Diabetes Research (DZD e.V., grant number 01G10925), the Deutsche Forschungsgemeinschaft (DFG, grant number

SO932-2 to SRS and Reinhart Koselleck support to NB); the European Commission under the project WAY (grant number 288551 to SRS and NB); the Volkswagenstiftung (VW) and the Baden-Württemberg Stiftung, Germany.

References

Ang, K.K., Guan, C., Chua, K.S.G., Ang, B.T., Kuah, C.W.K., Wang, C., Phua, K.S., Chin, Z.Y., Zhang, H., 2009. A clinical study of motor imagery-based brain-computer interface for upper limb robotic rehabilitation. Proceedings of the 31st Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Minneapolis, MN, pp. 5981-5984.

Ang, K.K., Guan, C., Chua, K.S., Ang, B.T., Kuah, C., Wang, C., et al., 2010. Clinical study ofneurorehabilitation in stroke using EEG-based motor imagery brain-computer interface with robotic feedback. Conf Proc IEEE Eng Med. Biol. Soc. 1, pp. 5549-5552

Ang, K.K., Chua, K.S., Phua, K.S., Wang, C., Chin, Z.Y., Kuah, C.W., Low, W., Guan, C., 2014. A randomized controlled trial of EEG-based motor imagery brain-computer interface robotic rehabilitation for stroke. Clin. EEG Neurosci. http://dx.doi.org/10.1177/ 1550059414522229 (in press).

Basmajian, J.V., 1981. Biofeedback in rehabilitation: a review of principles and practices. Arch. Phys. Med. Rehabil. 62,469-475.

Basmajian, J.V., Gowland, C., Brandstater, M.E., Swanson, L., Trotter, J., 1982. EMG feedback treatment of upper limb in hemiplegic stroke patients: a pilot study. Arch. Phys. Med. Rehabil. 63, 613-616.

Birbaumer, N., Cohen, L.G., 2007. Brain-computer interfaces: communication and restoration of movement in paralysis. J. Physiol. 579,621-636.

Birbaumer, N., Ghanayim, N., Hinterberger, T., Iversen, I., Kotchoubey, B., Kubler, A., Perelmouter, J., Taub, E., Flor, H., 1999. A spelling device for the paralysed. Nature 398, 297-298.

Birbaumer, N., Gallegos-Ayala, G., Wildgruber, M., Silvoni, S., Soekadar, S.R., 2014. Direct brain control and communication in paralysis. Brain Topogr. 27,4-11.

Birbeck, G.L., Hanna, M.G., Griggs, R.C., 2014. Global opportunities and challenges for clinical neuroscience. JAMA 311,1609-1610.

Bolwig, T.G., 2014. Neuroimaging and electroconvulsive therapy: a review. J. ECT 30 (2), 138-142.

Brasil, F., Curado, M.R., Witkowski, M., Garcia, E., Broetz, D., Birbaumer, N., Soekadar, S.R., 2012. MEP predicts motor recovery in chronic stroke patients undergoing 4-weeks of daily physical therapy. Human Brain Mapping Annual Meeting, Beijing. June 10-14.

Broetz, D., Braun, C., Weber, C., Soekadar, S.R., Caria, A., Birbaumer, N., 2010. Combination of brain-computer interface training and goal-directed physical therapy in chronic stroke: a case report. Neurorehabil. Neural Repair 24, 674-679.

Broetz, D., Del Grosso, N.A., Rea, M., Ramos-Murguialday, A., Soekadar, S.R., Birbaumer, N., 2014. A new hand assessment instrument for severely affected stroke patients. NeuroRehabilitation 34,409-427.

Buch, E., Weber, C., Cohen, L.G., Braun, C., Dimyan, M.A., Ard, T., Mellinger, J., Caria, A., Soekadar, S.R., Fourkas, A., Birbaumer, N., 2008. Think to move: a neuromagnetic brain-computer interface (BCl) system for chronic stroke. Stroke 39,910-917.

Buch, E.R., Modir, Shanechi A., Fourkas, A.D., Weber, C., Birbaumer, N., Cohen, L.G., 2012. Parietofrontal integrity determines neural modulation associated with grasping imagery after stroke. Brain 135, 596-614.

Buma, F., Kwakkel, G., Ramsey, N., 2013. Understanding upper limb recovery after stroke. Restor. Neurol. Neurosci. 31, 707-722.

Bundy, D.T., Wronkiewicz, M., Sharma, M., Moran, D.W., Corbetta, M., Leuthardt, E.C., 2012. Using ipsilateral motor signals in the unaffected cerebral hemisphere as a signal platform for brain-computer interfaces in hemiplegic stroke survivors. J. Neural Eng. 9, 036011. http://dx.doi.org/10.1088/1741-2560/9/3/036011.

Burke, E., Cramer, S.C., 2013. Biomarkers and predictors of restorative therapy effects after stroke. Curr. Neurol. Neurosci. Rep. 13, 329.

Calautti, C., Jones, P.S., Naccarato, M., Sharma, N., Day, D.J., Bullmore, E.T., Warburton, E.A., Baron, J.C., 2010. The relationship between motor deficit and primary motor cortex hemispheric activation balance after stroke: longitudinal fMRl study. J. Neurol. Neurosurg. Psychiatry 81, 788-792.

Caria, A., Weber, C., Brotz, D., Ramos, A., Ticini, L.F., Gharabaghi, A., Braun, C., Birbaumer, N., 2011. Chronic stroke recovery after combined BCl training and physiotherapy: a case report. J. Psychophysiol. 48, 578-582.

Carmena, J.M., 2013. Advances in neuroprosthetic learning and control. PLoS Biol. 11, e1001561. http://dx.doi.org/10.1371/journal.pbio.1001561.

Carmena, J.M., Lebedev, M.A., Crist, R.E., O'Doherty, J.E., Santucci, D.M., Dimitrov, D.F., Patil, P.G., Henriquez, C.S., Nicolelis, MAL., 2003. Learning to control a brain-machine interface for reaching and grasping by primates. PLoS Biol. 1, E42. http://dx.doi.org/10. 1371 /journal.pbio.0000042.

Chollet, F., Tardy, J., Albucher, J.F., Thalamas, C., Berard, E., Lamy, C., Bejot, Y., Deltour, S., Jaillard, A., Niclot, P., Guillon, B., Moulin, T., Marque, P., Pariente, J., Arnaud, C., Loubinoux, I., 2011. Fluoxetine for motor recovery after acute ischaemic stroke (FLAME): a randomised placebo-controlled trial. Lancet Neurol. 10,123-130.

Chollet, F., Cramer, S.C., Stinear, C., Kappelle, L.J., Baron, J.C., Weiller, C., Azouvi, P., Hommel, M., Sabatini, U., Moulin, T., Tardy, J., Valenti, M., Montgomery, S., Adams Jr, H., 2014. Pharmacological therapies in post stroke recovery: recommendations for future clinical trials. J. Neurol. 261,1461-1468.

Cirstea, M.C., Levin, M.F., 2000. Compensatory strategies for reaching in stroke. Brain 123, 940-953.

ARTICLE IN PRESS

S.R. Soekadar et al. / Neurobiology of Disease xxx (2014) xxx-xxx 7

Collinger, J.L., Wodlinger, B., Downey, J.E., Wang, W., Tyler-Kabara, E.C., Weber, D.J., McMorland, A.J., Velliste, M., Boninger, M.L., Schwartz, A.B., 2013. High-performance neuroprosthetic control by an individual with tetraplegia. Lancet 381,557-564.

Daly, J.J., Wolpaw, J.R., 2008. Brain-computer interfaces in neurological rehabilitation. Lancet Neurol. 7,1032-1043. http://dx.doi.org/10.1016/S1474-4422(08)70223-0.

Daly, J.J., Cheng, R., Rogers, J., Litinas, K., Hrovat, K., Dohring, M., 2009. Feasibility of a new application of noninvasive brain computer interface (BCl): a case study of training for recovery of volitional motor control after stroke. J. Neurol. Phys. Ther. 33,203-211.

Dayan, E., Censor, N., Buch, E.R., Sandrini, M., Cohen, L.G., 2013. Noninvasive brain stimulation: from physiology to network dynamics and back Nat. Neurosci. 16, 838-844.

Dobkin, B.H., 2007. Brain-computer interface technology as a tool to augment plasticity and outcomes for neurological rehabilitation. J. Physiol. 579, 637-642.

Ethier, C., Oby, E.R., Bauman, M.J., Miller, L.E., 2012. Restoration of grasp following paralysis through brain-controlled stimulation of muscles. Nature 485, 368-371. http://dx.doi.org/10.1038/nature10987.

Farwell, La, Donchin, E., 1988. Talking off the top of your head: toward a mental prosthesis utilizing event-related brain potentials. Electroencephalogr. Clin. Neurophysiol. 70, 510-523.

Feeney, D.M., Baron, J.C., 1986. Diaschisis. Stroke 17, 817-830.

Feigin, V.L., Forouzanfar, M.H., Krishnamurthi, R., Mensah, G.A., Connor, M., Bennett, D.A., Moran, A.E., Sacco, R.L., Anderson, L., Truelsen, T., O'Donnell, M., Venketasubramanian, N., Barker-Collo, S., Lawes, C.M., Wang, W., Shinohara, Y., Witt, E., Ezzati, M., Naghavi, M., Murray, C., 2014. Global and regional burden of stroke during 1990-2010: findings from the global burden of disease study 2010. Global burden of diseases, injuries, and risk factors study 2010 (GbD 2010) and the GBD stroke experts group. Lancet 383, 245-254.

Flint, R.D., Ethier, C., Oby, E.R., Miller, L.E., Slutzky, M.W., 2012a. Local field potentials allow accurate decoding of muscle activity. J. Neurophysiol. 108,18-24. http://dx. doi.org/10.1152/jn.00832.2011.

Flint, R.D., Lindberg, E.W., Jordan, L.R., Miller, L.E., Slutzky, M.W., 2012b. Accurate decoding of reaching movements from field potentials in the absence of spikes. J. Neural Eng. 9, 046006. http://dx.doi.org/10.1088/1741-2560/9/4/046006.

Flint, R.D., Wright, Z.A., Scheid, M.R., Slutzky, M.W., 2013. Long term, stable brain machine interface performance using local field potentials and multiunit spikes. J. Neural Eng. 10, 056005. http://dx.doi.org/10.1088/1741-2560/10/5Z056005.

Flint, R.D., Wang, P.T., Wright, Z.A., King, C.E., Krucoff, M.O., Schuele, S.U., Rosenow, J.M., Hsu, F.P., Liu, C.Y., Lin, J.J., Sazgar, M., Millett, D.E., Shaw, S.J., Nenadic, Z., Do, A.H., Slutzky, M.W., 2014. Extracting kinetic information from human motor cortical signals. Neuroimage 101,695-703.

Galea, J.M., Celnik, P., 2009. Brain polarization enhances the formation and retention of motor memories. J. Neurophysiol. 102, 294-301.

Guggenmos, D.J., Azin, M., Barbay, S., Mahnken, J.D., Dunham, C., Mohseni, P., Nudo, R.J., 2013. Restoration of function after brain damage using a neural prosthesis. PNAS 110, 21177-21182.

Hackett, M.L., Köhler, S., O'Brien, J.T., Mead, G.E., 2014. Neuropsychiatric outcomes of stroke. Lancet Neurol. 13, 525-534.

Hao, Z., Wang, D., Zeng, Y., Liu, M., 2013. Repetitive transcranial magnetic stimulation for improving function after stroke. Cochrane Database Syst. Rev. 5, CD008862.

Heck C.N., King-Stephens, D., Massey, A.D., Nair, D.R., Jobst, B.C., Barkley, G.L., Salanova, V., Cole, A.J., Smith, M.C., Gwinn, R.P., et al., 2014. Two-year seizure reduction in adults with medically intractable partial onset epilepsy treated with responsive neurostimulation: final results of the RNS system pivotal trial. Epilepsia 55, 432-441. http://dx.doi.org/10.1111 /epi.12534.

Hochberg, L.R., Serruya, M.D., Friehs, G.M., Mukand, J.A., Saleh, M., Caplan, A.H., Branner, A., Chen, D., Penn, R.D., Donoghue, J.P., 2006. Neuronal ensemble control of prosthetic devices by a human with tetraplegia. Nature 442,164-171. http://dx.doi.org/10. 1038/nature04970.

Hochberg, L.R., Bacher, D., Jarosiewicz, B., Masse, Nicolas, Y., Simeral, J.D., Vogel, J., Haddadin, S., Liu, J., Cash, S.S., van der Smagt, P., Donoghue, J.P., 2012. Reach and grasp by people with tetraplegia using a neurally controlled robotic arm. Nature 485,372-375. http://dx.doi.org/10.1038/nature11076.

Hummel, F.C., Voller, B., Celnik P., Floel, A., Giraux, P., Gerloff, C., Cohen, L.G., 2006. Effects of brain polarization on reaction times and pinch force in chronic stroke. BMC Neurosci. 7, 73.

Hwang, E.J., Andersen, R.A., 2009. Brain control of movement execution onset using local field potentials in posterior parietal cortex. J. Neurosci. 29,14363-14370. http://dx. doi.org/10.1523/JNEUROSCl.2081-09.2009.

Jung, S.H., Kim, Y.K., Kim, S.E., Paik, N.J., 2012. Prediction of motor function recovery after subcortical stroke: case series of activation PET and TMS studies. Ann. Rehabil. Med. 36,501-511.

Khedr, E.M., Etraby, A.E., Hemeda, M., Nasef, A.M., Razek A.A., 2010. Long-term effect of repetitive transcranial magnetic stimulation on motor function recovery after acute ischemic stroke. Acta Neurol. Scand. 121, 30-37. http://dx.doi.org/10.1111/j.1600-0404.2009.01195.x.

Koralek, A.C., Jin, X., Long ll, J.D., Costa, R.M., Carmena, J.M., 2012. Corticostriatal plasticity is necessary for learning intentional neuroprosthetic skills. Nature 483, 331-335.

Lang, C.E., Wagner, J.M., Edwards, D.F., Sahrmann, S.A., Dromerick A.W., 2006. Recovery of grasp versus reach in people with hemiparesis poststroke. Neurorehabil. Neural Repair 20, 444-454.

Langhorne, P., Bernhardt, J., Kwakkel, G., 2011. Stroke rehabilitation. Lancet 377, 1693-1702.

Leuthardt, E.C., Schalk, G., Wolpaw, J.R., Ojemann, J.G., Moran, D.W., 2004. A brain-computer interface using electrocorticographic signals in humans. J. Neural Eng. 1, 63-71. http://dx.doi.org/10.1088/1741-2560/1/2/001.

Liew, S.L., Santarnecchi, E., Buch, E.R., Cohen, L.G., 2014. Non-invasive brain stimulation in neurorehabilitation: local and distant effects for motor recovery. Front. Hum. Neurosci. 8, 378.

Linden, D.E., Habes, I., Johnston, S.J., Linden, S., Tatineni, R., Subramanian, L., Sorger, B., Healy, D., Goebel, R., 2012. Real-time self-regulation of emotion networks in patients with depression. PLoS One 7, e38115.

Lopez, A.D., Mathers, C.D., Ezzati, M., Jamison, D.T., Murray, C.J., 2006. Global and regional burden of disease and risk factors, 2001: systematic analysis of population health data. Lancet 367,1747-1757.

Lubar, J.F., Shouse, M.N., 1976. EEG and behavioral changes in a hyperkinetic child concurrent with training of the sensorimotor rhythm (SMR): a preliminary report. Biofeedback Self Regul. 1, 293-306.

Marshall, L., Mölle, M., Hallschmid, M., Born, J, 2004. Transcranial direct current stimulation during sleep improves declarative memory. J. Neurosci. 24, 9985-9992.

McFarland, D.J., Neat, G.W., Read, R.F., Wolpaw, J.R., 1993. An EEG-based method for graded cursor control. Psychobiology 21, 77-81.

McFarland, D.J., Krusienski, D.J., Wolpaw, J.R., 2006. Brain-computer interface signal processing at the Wadsworth Center: mu and sensorimotor beta rhythms. Prog. Brain Res. 159,411-419.

McGie, S.C., Zariffa, J., Popovic, M.R., Nagai, M.K., 2014. Short-term neuroplastic effects of brain-controlled and muscle-controlled electrical stimulation. Neuromodulation http://dx.doi.org/10.1111/ner.12185 (in press).

Mead, G.E., Hsieh, C.F., Hackett, M., 2013. Selective serotonin reuptake inhibitors for stroke recovery. JAMA 310,1066-1067.

Mehring, C., Nawrot, M.P., de Oliveira, S.C., Vaadia, E., Schulze-Bonhage, A., Aertsen, A., Ball, T., 2004. Comparing information about arm movement direction in single channels of local and epicortical field potentials from monkey and human motor cortex. J. Physiol. Paris98,498-506. http://dx.doi.org/10.1016/jophysparis.2005.09.016.

Mihara, M., Hattori, N., Hatakenaka, M., Yagura, H., Kawano, T., Hino, T., Miyai, I., 2013. Near-infrared spectroscopy-mediated neurofeedback enhances efficacy of motor imagery-based training in poststroke victims: a pilot study. Stroke 44 (4), 1091-1098.

Millän, J.D., Rupp, R., Müller-Putz, G.R., Murray-Smith, R., Giugliemma, C., Tangermann, M., Vidaurre, C., Cincotti, F., Kübler, A., Leeb, R., Neuper, C., Müller, K.R., Mattia, D., 2010. Combining brain-computer interfaces and assistive technologies: state-of-the-art and challenges. Front. Neurosci. 4, 161.

Monastra, V.J., Lynn, S., Linden, M., Lubar, J.F., Gruzelier, J., LaVaque, T.J., 2005. Electroen-cephalographic biofeedback in the treatment of attention-deficit/hyperactivity disorder. Appl. Psychophysiol. Biofeedback 30,95-114.

Moritz, C.T., Perlmutter, S.I., Fetz, E.E., 2008. Direct control of paralysed muscles by cortical neurons. Nature 456, 639-642. http://dx.doi.org/10.1038/nature07418.

Mukaino, M., Ono, T., Shindo, K., Fujiwara, T., Ota, T., Kimura, A., Liu, M., Ushiba, J., 2014. Efficacy of brain-computer interface-driven neuromuscular electrical stimulation for chronic paresis after stroke. J. Rehabil. Med. 46, 378-382.

Murthy, V.N., Fetz, E.E., 1996. Synchronization of neurons during local field potential oscillations in sensorimotor cortex of awake monkeys. J. Neurophysiol. 76,3968-3982.

Nishimura, Y., Perlmutter, S.I., Fetz, E.E., 2013. Restoration of upper limb movement via artificial corticospinal and musculospinal connections in a monkey with spinal cord injury. Front. Neural Circ. 7, 57.

Nudo, R.J.J., Milliken, G.W.W., 1996. Reorganization of movement representations in primary motor cortex following focal ischemic infarcts in adult squirrel monkeys. J. Neurophysiol. 75, 2144-2149.

Ono, T., Shindo, K., Kawashima, K., Ota, N., Ito, M., Ota, T., Mukaino, M., Fujiwara, T., Kimura, A., Liu, M., Ushiba, J., 2014. Brain-computer interface with somatosensory feedback improves functional recovery from severe hemiplegia due to chronic stroke. Front. Neuroeng. 7,19.

Pfurtscheller, G., Müller, G.R., Pfurtscheller, J., Gerner, H.J., Rupp, R., 2003. Thought-control of functional electrical stimulation to restore hand grasp in a patient with tetraplegia. Neurosci. Lett. 351,33-36.

Pfurtscheller, G., Brunner, C., Schlögl, A., Lopes da Silva, F.H., 2006. Mu rhythm (de) synchronization and EEG single-trial classification of different motor imagery tasks. Neuroimage 31, 153-159.

Pirulli, C., Fertonani, A., Miniussi, C., 2013. The role of timing in the induction of neuromodulation in perceptual learning by transcranial electric stimulation. Brain Stimul. 6, 683-689.

Platz, T., Kim, I.H., Engel, U., Kieselbach, A., Mauritz, K.-H., 2002. Brain activation pattern as assessed with multi-modal EEG analysis predict motor recovery among stroke patients with mild arm paresis who receive the Arm Ability Training. Restor. Neurol. Neurosci. 20, 21-35.

Pohlmeyer, E.A., Oby, E.R., Perreault, E.J., Solla, S.A., Kilgore, K.L., Kirsch, R.F., Miller, L.E., 2009. Toward the restoration of hand use to a paralyzed monkey: brain-controlled functional electrical stimulation of forearm muscles. PLoS ONE 4, e5924. http://dx. doi.org/10.1371 /journal.pone.0005924.

Prasad, G., Herman, P., Coyle, D., McDonough, S., Crosbie, J., 2010. Applying a brain-computer interface to support motor imagery practice in people with stroke for upper limb recovery: a feasibility study. J. Neuroeng. Rehabil. 7(1), 60.

Ramos-Murguialday, A., Broetz, D., Rea, M., Läer, L., Yilmaz, O., Brasil, F.L., Liberati, G., Curado, M.R., Garcia-Cossio, E., Vyziotis, A., Cho, W., Agostini, M., Soares, E., Soekadar, S.R., Caria, A., Cohen, L.G., Birbaumer, N., 2013. Brain-machine interface in chronic stroke rehabilitation: a controlled study. Ann. Neurol. 74,100-108.

Rea, M., Rana, M., Lugato, N., Terekhin, P., Gizzi, L., Brötz, D., Fallgatter, A., Birbaumer, N., Sitaram, R., Caria, A., 2014. Lower limb movement preparation in chronic stroke: A pilot study toward an fNIRS-BCI for gait rehabilitation, Neurorehabil. Neural Repair 28, 564-575.

Reis, J., Robertson, E., Krakauer, J.W., Rothwell, J., Marshall, L., Gerloff, C., Wassermann, E., Pascual-Leone, A., Hummel, F., Celnik, P.A., Classen, J., Floel, A., Ziemann, U., Paulus,

ARTICLE IN PRESS

8 S.R. Soekadar et al. / Neurobiology of Disease xxx (2014) xxx-xxx

W., Siebner, H.R., Born, J., Cohen, L.G., 2008. Consensus: "can tDCS and TMS enhance motor learning and memory formation?". Brain Stimul. 1, 363-369.

Rozelle, G.R., Budzynski, T.H., 1995. Neurotherapy for stroke rehabilitation: a single case study. Biofeedback Self Regul. 20,211-228.

Ruiz, S., Buyukturkoglu, K., Rana, M., Birbaumer, N., Sitaram, R., 2014. Real-time fMRI brain computer interfaces: self-regulation of single brain regions to networks. Biol. Psychol. 95, 4-20.

Sakurada, T., Kawase, T., Takano, K., Komatsu, T., Kansaku, K., 2013. A BMI-based occupational therapy assist suit: asynchronous control by SSVEP. Front. Neurosci. 7,172.

Sanes, J.N., Donoghue, J.P., 1993. Oscillations in local field potentials of the primate motor cortex during voluntary movement. PNAS 90,4470-4474.

Schalk G., Kubanek J., Miller, K.J., Anderson, N.R., Leuthardt, E.C., Ojemann, J.G., Limbrick D., Moran, D., Gerhardt, L.A., Wolpaw, J.R., 2007. Decoding two-dimensional movement trajectories using electrocorticographic signals in humans. J. Neural Eng. 4, 264-275. http://dx.doi.org/10.1088/1741-2560/473/012.

Schalk, G., Brunner, P., Gerhardt, L.A., Bischof, H., Wolpaw, J.R., 2008. Brain-computer interfaces (BCIs): detection instead of classification. J. Neurosci. Methods 167, 51-62. http://dx.doi.org/10.1016/jjneumeth.2007.08.010.

Sellers, E.W., Ryan, D.B., Hauser, C.K., 2014. Noninvasive brain-computer interface enables communication after brainstem stroke. Sci. Transl. Med. 6, 257re7.

Serruya, M.D., Hatsopoulos, N.G., Paninski, L., Fellows, M.R., Donoghue, J.P., 2002. Instant neural control of a movement signal. Nature 416,141-142. http://dx.doi.org/10. 1038/416141a.

Shaikhouni, A., Donoghue, J.P., Hochberg, L.R., 2013. Somatosensory responses in a human motor cortex. J. Neurophysiol. 109, 2192-2204.

Shepherd, G.M., 2013. Corticostriatal connectivity and its role in disease. Nat. Rev. Neurosci. 14, 278-291.

Shindo, K., Kawashima, K., Ushiba, J., Ota, N., Ito, M., Ota, T., Kimura, A., Liu, M.G., 2011. Effects of neurofeedback training with an electroencephalogram-based brain-computer interface for hand paralysis in patients with chronic stroke: a preliminary case series study. J. Rehabil. Med. 43 (10), 951-957.

Sirtori, V., Corbetta, D., Moja, L., Gatti, R., 2009. Constraint-induced movement therapy for upper extremities in stroke patients. Cochrane Database Syst. Rev. 4, CD004433.

Sitaram, R., Caria, A., Birbaumer, N., 2009. Hemodynamic brain-computer interfaces for communication and rehabilitation. Neural Netw. 22, 1320-1328.

Sitaram, R., Veit, R., Stevens, B., Caria, A., Gerloff, C., Birbaumer, N., Hummel, F., 2012. Acquired control of ventral premotor cortex activity by feedback training: an exploratory real-time FMRI and TMS study. Neurorehabil. Neural Repair 26, 256-265.

Slutzky, M.W., Jordan, L.R., Krieg, T., Chen, M., Mogul, D.J., Miller, L.E., 2010. Optimal spacing of surface electrode arrays for brain-machine interface applications. J. Neural Eng. 7, 26004. http://dx.doi.org/10.1088/1741 -2560/7/2/026004.

Slutzky, M.W., Jordan, L.R., Lindberg, E.W., Lindsay, K.E., Miller, L.E., 2011. Decoding the rat forelimb movement direction from epidural and intracortical field potentials. J. Neural Eng. 8, 036013. http://dx.doi.org/10.1088/1741-2560/8/3/036013.

Soekadar, S.R., Birbaumer, N., Cohen, L.G., 2011a. Brain-computer interfaces in the rehabilitation of stroke and neurotrauma. In: Cohen, L.G., Kansaku, K. (Eds.), Systems Neuroscience and Rehabilitation. Springer Berlin, Germany, pp. 3-18.

Soekadar, S.R., Witkowski, M., Mellinger, J., Ramos, A., Birbaumer, N., Cohen, L.G., 2011b. ERD-based online brain-machine interfaces (BMI) in the context of neurorehabilitation: optimizing BMI learning and performance. IEEE Trans. Neural. Syst. Rehabil. Eng. 19, 542-549.

Soekadar, S.R., Witkowski, M., Cossio, E.G., Birbaumer, N., Robinson, S.E., Cohen, L.G., 2013a. In vivo assessment of human brain oscillations during application of transcranial electric currents. Nat. Commun. 4, 2032.

Soekadar, S.R., Witkowski, M., Robinson, S.E., Birbaumer, N., 2013b. Combining electric brain stimulation and source-based brain-machine interface (BMI) training in neurorehabilitation of chronic stroke. J. Neurol. Sci. 333, e542.

Soekadar, S.R., Witkowski, M., Cossio, E.G., Birbaumer, N., Cohen, L.G., 2014. Learned EEG-based brain self-regulation of motor-related oscillations during application of transcranial electric brain stimulation: feasibility and limitations. Front. Behav. Neurosci. 8, 93.

Soekadar, S.R., Witkowski, M., Vitiello, N., Birbaumer, N., 2014. An EEG/EOG-based hybrid brain-neural computer interaction (BNCI) system to control an exoskeleton for the paralyzed hand. Biomed. Tech. (Berl.) http://dx.doi.org/10.1515/bmt-2014-0126 (in press-a).

Soekadar, S.R., Witkowski, M., Birbaumer, N., Cohen, L.G., 2014. Enhancing Hebbian learning to control brain oscillatory activity. Cereb. Cortex http://dx.doi.org/10. 1093/cercor/bhu043 (in press-b).

Stagg, C.J., Jayaram, G., Pastor, D., Kincses, Z.T., Matthews, P.M., Johansen-Berg, H., 2011. Polarity and timing-dependent effects of transcranial direct current stimulation in explicit motor learning. Neuropsychologia 49, 800-804.

Stark, E., Abeles, M., 2007. Predicting movement from multiunit activity. J. Neurosci. 27, 8387-8394. http://dx.doi.org/10.1523/JNEUROSCI. 1321-07.2007.

Sterman, M.B., Macdonald, L.R., 1978. Effects of central cortical EEG feedback training on incidence of poorly controlled seizures. Epilepsia 19,207-222.

Sterman, M.B., Wyrwicka, W., Howe, R., 1969. Behavioral and neurophysiological studies of the sensorimotor rhythm in the cat. Electroencephalogr. Clin. Neurophysiol. 27, 678-679.

Strehl, U., Leins, U., Goth, G., Klinger, C., Hinterberger, T., Birbaumer, N., 2006. Self-regulation of slow cortical potentials: a new treatment for children with attention-deficit/hyperactivity disorder. Pediatrics 118, e1530-e1540.

Sulzer, J., Sitaram, R., Blefari, M.L., Kollias, S., Birbaumer, N., Stephan, K.E., Luft, A., Gassert, R., 2013. Neurofeedback-mediated self-regulation of the dopaminergic midbrain. NeuroImage 83, 817-825.

Sun, H.Y., Xiang, Y., Yang, M.D., 2011. Neurological rehabilitation of stroke patients via motor imaginary-based brain-computer interface technology. Neural Regen. Res. 6 (28), 2198-2202.

Sung, W.H., Wang, C.P., Chou, C.L., Chen, Y.C., Chang, Y.C., Tsai, P.Y., 2013. Efficacy of coupling inhibitory and facilitatory repetitive transcranial magnetic stimulation to enhance motor recovery in hemiplegic stroke patients. Stroke 44, 1375-1382. http://dx.doi.org/10.1161/STROKEAHA111.000522.

Takahashi, M., Takeda, K., Otaka, Y., Osu, R., Hanakawa, T., Gouko, M., Ito, K., 2012. Event related desynchronization-modulated functional electrical stimulation system for stroke rehabilitation: a feasibility study. J. Neuroeng. Rehabil. 9,56.

Takeuchi, N., Chuma, T., Matsuo, Y., Watanabe, I., Ikoma, K., 2005. Repetitive transcranial magnetic stimulation of contralesional primary motor cortex improves hand function after stroke. Stroke 36, 2681-2686.

Taub, E., Uswatte, G., Pidikiti, R., 1999. Constraint-induced movement therapy: a new family of techniques with broad application to physical rehabilitation—a clinical review. J. Rehabil. Res. Dev. 36, 237-251.

Taub, E., Uswatte, G., Elbert, T., 2002. New treatments in neurorehabilitation founded on basic research. Nat. Rev. Neurosci. 3, 228-236.

Taylor, D.M., Tillery, S.I.H., Schwartz, A.B., 2002. Direct cortical control of 3D neuroprosthetic devices. Science 296 (5574), 1829-1832. http://dx.doi.org/10.1126/ science.1070291.

Teasell, R.W., Murie Fernandez, M., McIntyre, A., Mehta, S., 2014. Rethinking the continuum of stroke rehabilitation. Arch. Phys. Med. Rehabil. 95, 595-596.

Terroni, L., Amaro, E., Iosifescu, D.V., Tinone, G., Sato, J.R., Leite, C.C., Sobreiro, M.F., Lucia, M.C., Scaff, M., Fräguas, R., 2011. Stroke lesion in cortical neural circuits and post-stroke incidence of major depressive episode: a 4-month prospective study. World J. Biol. Psychiatry 12,539-548.

Värkuti, B., Guan, C., Pan, Y., Phua, K.S., Ang, K.K., Kuah, C.W., Chua, K., Ang, B.T., Birbaumer, N., Sitaram, R., 2013. Resting state changes in functional connectivity correlate with movement recovery for BCI and robot-assisted upper-extremity training after stroke. Neurorehabil. Neural Repair 27, 53-62.

Venkatakrishnan, A., Francisco, G.E., Contreras-Vidal, J.L., 2014. Applications of brain-machine interface systems in stroke recovery and rehabilitation. Curr. Phys. Med. Rehabil. Rep. 2, 93-105.

Volpato, C., Cavinato, M., Piccione, F., Garzon, M., Meneghello, F., Birbaumer, N., 2013. Transcranial direct current stimulation (tDCS) of Broca's area in chronic aphasia: a controlled outcome study. Behav. Brain Res. 247, 211-216. http://dx.doi.org/10. 1016/j.bbr.2013.03.029.

Wang, W., Collinger, J.L., Perez, M.A., Tyler-Kabara, E.C., Cohen, L.G., Birbaumer, N., Brose, S.W., Schwartz, A.B., Boninger, M.L., Weber, D.J., 2010. Neural interface technology for rehabilitation: exploiting and promoting neuroplasticity. Phys. Med. Rehabil. Clin. N. Am. 21,157-178.

Wang, W., Collinger, J.L., Degenhart, A.D., Tyler-Kabara, E.C., Schwartz, A.B., Moran, D.W., Weber, D.J., Wodlinger, B., Vinjamuri, R.K., Ashmore, R.C., et al., 2013. An electrocorticographic brain interface in an individual with tetraplegia. PLoS ONE 8, e55344. http://dx.doi.org/10.1371 /journal.pone.0055344.

Weiskopf, N., Veit, R., Erb, M., Mathiak, K., Grodd, W., Goebel, R., Birbaumer, N., 2003. Physiological self-regulation of regional brain activity using real-time functional magnetic resonance imaging (fMRI): methodology and exemplary data. Neurolmage 19, 577-586.

Wolf, S.L., Lecraw, D.E., Barton, L.A., Jann, B.B., 1989. Forced use of hemiplegic upper extremities to reverse the effect of learned nonuse among chronic stroke and head-injured patients. Exp. Neurol. 104,125-132.

Young, B.M., Nigogosyan, Z., Walton, L.M., Song, J., Nair, V.A., Grogan, S.W., Tyler, M.E., Edwards, D.F., Caldera, K., Sattin,J.A., Williams, J.C., Prabhakaran, V., 2014. Changes in functional brain organization and behavioral correlations after rehabilitative therapy using a brain-computer interface. Front. Neuroeng. 7, 26.

Zhuang, J., Truccolo, W., Vargas-Irwin, C., Donoghue, J.P., 2010. Decoding 3-D reach and grasp kinematics from high-frequency local field potentials in primate primary motor cortex. IEEE Trans. Biomed. Eng. 57,1774-1784. http://dx.doi.org/10.1109/ TBME.2010.2047015.

Further reading

Link to video 1: https://www.youtube.com/watch?v=ogBX18maUiM.Link to video 2: https://www.youtube.com/watch?v=76lIQtE8oDY.