Provided for non-commercial research and educational use only. Not for reproduction, distribution or commercial use.

This chapter was originally published in the book International Review of Neurobiology, Vol. 86, published by Elsevier, and the attached copy is provided by Elsevier for the author's benefit and for the benefit of the author's institution, for noncommercial research and educational use including without limitation use in instruction at your institution, sending it to specific colleagues who know you, and providing a copy to your institutions administrator.

All other uses, reproduction and distribution, including without limitation commercial reprints, selling or licensing copies or access, or posting on open internet sites, your personal or institutions website or repository, are prohibited. For exceptions, permission may be sought for such use through Elsevier's permissions site at: http://www.elsevier.com/locate/permissionusematerial From: Niels Birbaumer, Ander Ramos Murguialday, Cornelia Weber and Pedro Montoya, Neurofeedback and BrainComputer Interface: Clinical Applications. In Luca Rossini, Dario Izzo and Leopold Summerer, editors: International Review of Neurobiology, Vol. 86, Burlington: Academic Press, 2009, pp. 107-117. ISBN: 978-0-12-374821-8 Copyright 2009 Elsevier Inc. Academic Press.

Author's personal copy

NEUROFEEDBACK AND BRAINCOMPUTER INTERFACE: CLINICAL APPLICATIONS

Niels Birbaumer,*,y Ander Ramos Murguialday,*,z Cornelia Weber,* and Pedro Montoya}
*Institute of Medical Psychology and Behavioral Neurobiology, University of Tuebingen, Tuebingen, Germany y Ospedale San CamilloIRCCS, Istituto di Ricovero e Cura a Carattere Scientifico, Venezia, Lido, Italy z Fatronik Foundation, San Sebastian, Spain } Department of Psychology, Universidad Illes Baleares, Palma de Mallorca, Spain

I. II. III. IV. V.

Introduction Functional Magnetic Resonance Imaging: f MRI-BMI BMI in Locked-in Syndrome BMI in Stroke and Spinal Cord Injury Conclusion References

Most of the research devoted to BMI development consists of methodological studies comparing diVerent online mathematical algorithms, ranging from simple linear discriminant analysis (LDA) (Dornhege et al., 2007) to nonlinear articial neural networks (ANNs) or support vector machine (SVM) classication. Single cell spiking for the reconstruction of hand movements requires diVerent statistical solutions than electroencephalography (EEG)-rhythm classication for communication. In general, the algorithm for BMI applications is computationally simple and diVerences in classication accuracy between algorithms used for a particular purpose are small. Only a very limited number of clinical studies with neurological patients are available, most of them single case studies. The clinical target populations for BMI-treatment consist primarily of patients with amyotrophic lateral sclerosis (ALS) and severe CNS damage including spinal cord injuries and stroke resulting in substantial decits in communication and motor function. However, an extensive body of literature started in the 1970s using neurofeedback training. Such training implemented to control various EEG-measures provided solid evidence of positive eVects in patients with otherwise pharmacologically intractable epilepsy, attention decit disorder, and hyperactivity ADHD. More recently, the successful introduction and testing of real-time f MRI and a NIRS-BMI opened an exciting eld of interest in patients with psychopathological conditions.
INTERNATIONAL REVIEW OF NEUROBIOLOGY, VOL. 86 DOI: 10.1016/S0074-7742(09)86008-X

107

Copyright 2009, Elsevier Inc. All rights reserved. 0074-7742/09 $35.00

Author's personal copy

108

BIRBAUMER et al.

I. Introduction

Most clinical applications of BMI-research rest on the tradition of neurofeedback and biofeedback, both consequences of technological achievements in rapid computer analysis of EEG patterns that allow online feedback and reward of diVerent types of neuroelectric activity (Elbert et al., 1984). BMIs aimed at restoration of movement, however, were built in the tradition of tuning functions of sensori-motor neurons representing diVerent directions of movements (Georgopoulos et al., 2007). Neurofeedback allowed, for the rst time, voluntary self-regulation of brain activity through feedback and reward. Expectancies ran high and many premature announcements of clinical success based on single case studies or uncontrolled observations discredited the eld early on. In the 1970s Millers demonstrations of operant control of autonomic (and CNS) functions (Miller, 1969) in curarized rats, supposedly proving voluntary operant regulation of many bodily functions excluding mediation of the motor system through curarization, turned out to be diYcult to replicate (Dworkin and Miller, 1986). Together with the clinical overstatements in the eld of biofeedback, this historic incident virtually halted funding from public sources and blocked large controlled clinical studies despite some indications of its eYciency. However, more recent studies suggested that some patients with drug-resistant epilepsy (mostly with secondarily generalized seizures) experienced a reduction in the number of ictal events during and after training consistent of self-regulation of slow cortical potentials (SCPs) (Kotchoubey et al., 2001; Rockstroh et al., 1993), an eVect also reported using biofeedback of skin conductance responses (GSR) (Nagai et al., 2004). Nagai et al. showed that learned increase in autonomic arousal through reduction of skin conductance decreased negative SCPs at the cortical level and thus increased seizure thresholds conrming earlier reports (Birbaumer et al., 1990; Kotchoubey et al., 2001; Rockstroh et al., 1993). In those studies with training and visual feedback of positive SCPs in focal epilepsies, some patients achieved virtually 100% accuracy in the control of SCPs after extensive training of 3050 sessions, thus paving the way for application to BMIs for communication. Still, well-controlled trials with larger samples of epileptic patients have not been implemented. Another promising line of neurofeedback in neurology is the self-regulation of SCPs and mu-rhythm (also called sensori-motor-rhythm, SMR) in attention decit disorder and hyperactivity (ADHD). SMR occurs over the sensorimotor rolandic brain regions with a frequency of 815 Hz indicating motor quiescence and a functionally inhibitory mode of the thalamocortical loops (Sterman and Clemente, 1962a). Motor imagery or motor action desynchronizes SMR (eventrelated desynchronization, ERD). Well-controlled studies with relatively small

Author's personal copy

NEUROFEEDBACK AND BCI

109

samples of ADHD children showed potential, pointing to lasting eVects on attention, and vigilance comparable to those achieved through pharmacological treatment with stimulants (Fuchs et al., 2003; Strehl et al., 2006). All in all, these pioneering studies underlined the possibility to control human electrocortical activity and to modify motor and cognitive functions in health and disease.

II. Functional Magnetic Resonance Imaging: f MRI-BMI

Near-infrared-spectroscopy (NIRS) measuring changes in oxygenation and in deoxygenation of the cortical surface is a relatively cheap noninvasive technology whose regulation can be learned within a few training sessions with contingent feedback only (Sitaram et al., 2007). Sitaram et al. trained healthy human subjects successfully to maximize the diVerence between right and left sensorimotor regions. The regulation of the blood oxygenation level dependent (BOLD) response with real-time fMRI (rt-fMRI) is another development in BMI research (Weiskopf et al., 2005) (see Fig. 1). In contrast to all other noninvasive BMI measures, regulation of circumscribed cortical and subcortical structures is possible. Several experimental studies, mostly with young healthy volunteers, revealed an amazing anatomical resolution in the characterization of the cortical region to be trained and a good correlation of these changes with behavioral changes. For example, regulation of premotor and motor areas leads to changes in motor response speed ( Weiskopf et al., 2005), of anterior cingulated regions to downregulation of pain (DeCharms et al., 2005), of parahippocampal areas to changes in explicit memory performance ( Weiskopf et al., 2005) and of the anterior insula to changes in the valence of negative emotional slides without aVecting neutral or positive emotions (Caria et al., 2007). Healthy subjects are able to increase and decrease BOLD activity in a region of interest within one to three 1-h training sessions: usually they receive positive visual feedback within a second after the BOLD change (which itself has a latency of 23 s to the neural response). Experiments manipulating the connectivity between diVerent brain areas and real-time control of selected metabolic substances in specic brain regions using magnetic resonance spectroscopy-feedback are underway. A recent experiment (Caria et al., 2007) trained healthy subjects and criminal psychopaths to increase BOLD in the right insula. Before and after training, usually one to three 1-h sessions, aversive and neutral slides were presented. One control group received sham feedback and a second control group received training of emotional imagery. Only in the experimental group and for neutral slides learned BOLD increase in the insula increased subjective aversiveness of

Author's personal copy

110

BIRBAUMER et al.

Overview of fMRI-BCI system

Raw images Signal analysis Signal feedback

Participant

BCI program

Preprocessed brain activity

Visual feedback and reward computation Pattern classification (support vector machine) Performance measures

FIG. 1. Experimental condition for f MRI-BCI and BOLD-neurofeedback (see text for explanation).

picture stimuli. This study conrms the anatomical specicity of the f MRI-BMI eVects and underscores the critical importance of contingent feedback and reward. MR-technology use is expensive and applications in large clinical groups may not be feasible but represents a powerful tool to explore the mechanisms underlying BMI eVects and brainbehaviorpathology relationships in emotional disorders such as psychopathy and substance abuse as well as in other neuropsychiatric conditions.

III. BMI in Locked-in Syndrome

Patients with progressive motor neuron disease, particular amyotrophic later al sclerosis (ALS), GuillainBarre Syndrome, and subcortical stroke, as well as patients with traumatic brain damage in vegetative state (Kotchoubey, 2005) may suVer from locked-in syndrome (LIS) or total locked-in syndrome (TLIS). LIS is dened as complete paralysis with one or a few voluntary functions left (usually

Author's personal copy

NEUROFEEDBACK AND BCI

111
10 8 6 4 2

R Sess 1

R Sess 2

R Sess 3

FIG. 2. EVect of reward and feedback on BOLD-response of right insula (one person as representative example). Each sessions duration 20 min. Large increase of right insula BOLD after third session.

small eye movements). TLIS consists of complete cessation of volitional control of all voluntary somatic-motor functions. Both, LIS and TLIS show intact auditory and tactile perception and intact cognitive functions, usually measured with event-related brain potentials (ERP, Kotchoubey, 2005) or f MRI (Owen et al., 2006). Visual perception is also frequently compromised through paralysis of eye muscles. Therefore, BMIs using the auditory or tactile modality are mandatory for use in TLIS patients. Since the rst report (Birbaumer et al., 1999) of two LIS patients with ALS selecting letters from computer-presented letter strings using learned voluntary decrease of SCPs, several papers with small samples of ALS patients have appeared that demonstrate BMI controlled communication in LIS and advanced stages of ALS. In a thorough review of the literature it was proposed that BMIs using P300 ERPs (Farwell and Donchin, 1988; Sellers and Donchin, 2006), SCPs (Birbaumer et al., 1999) (see Fig. 3), and SMR-control (Pfurtscheller et al., 2005) could provide slow but eVective verbal communication in all stages of ALS, except the TLIS. It is of interest that in two patients with TLIS, not even an invasive BMI controlled from epidural electrodes at left frontal sites improved their ability to communicate (unpublished data, available from the authors). Only one study (Naito et al., 2007) reported more optimistic results from a NIRS-based BMI in 17 patients with TLIS. Patients were trained to respond with an increase in blood oxygenation (yes) or decrease in oxygenation (no) to various questions displayed on a computer screen. Using an elaborate oV-line classication method, a separation of yes and no of 70% correct was reported in seven out of 17 patients with TLIS. One weakness of this study is the relative lack of quantication and denition of clinical criterions used for the TLIS patients. It remains to be determined whether BMIs using EEG, electrocorticogram (ECoG), or NIRS allow voluntary brain responses and communication in TLIS. One possible explanation for the failure to replicate operant control of autonomic functions in the curarized rat (Dworkin and Miller, 1986; Miller, 1969) and for the lack of learned brain regulation with BMI in TLIS is that goal directed and voluntary

Author's personal copy

112

BIRBAUMER et al.

FIG. 3. Locked-in patient with amyotrophic lateral sclerosis using a braincomputer interface (BCI) for spelling. Top: Slow cortical potentials (SCPs) during one trial: a voluntary produced positive deection of the SCP splits the letter string on the screen below in half, if the desired letter is still among the letter string, a new positive SCP splits it again, and so forth until the desired letter is selected. Below: Screen with feedback cursor (not visible), letter string, from which the desired letter has to be selected (lowest box) and already spelled letters above the letter string.

thought processes may over time extinguish in the absence of reinforcement contingencies, a hypothesis worthwhile testing in the future (Birbaumer and Cohen, 2007). If this hypothesis is true a transfer of training success with a BMI from the LIS to the TLIS should be possible.

IV. BMI in Stroke and Spinal Cord Injury

Experimentation with nonhuman primates suggests that intentional goaldirected movements of the upper limbs can be reconstructed and transmitted to external manipulandum or robotic devices controlled from a relatively small number of microelectrodes implanted into movement-relevant brain areas after some training, opening the door for the development of braincomputer interfaces (BCIs) or brainmachine interfaces (BMIs) in humans. While noninvasive BMIs using electroencephalographic recordings EEG or ERPs in healthy individuals and patients with ALS or stroke, can transmit up to 80 bits/min of information, the use of BMIsinvasive or noninvasivein severely or totally paralyzed patients have met some unforeseen diYculties.

Author's personal copy

NEUROFEEDBACK AND BCI

113

In 2003, Pfurtscheller et al. reported a tetraplegic patient who, after extensive training to increase and decrease central mu-rhythms, was able to control an electrostimulation device (FES) applied to hand muscles (Pfurtscheller et al., 2000). The patient was able to grasp a glass and bring it to his mouth after he had learned with feedback and reward over a period of 4 months to regulate his mu-rhythm. Hochberg et al. (2006) implanted a 96-microelectrode array into the hand region of the motor cortex of another tetraplegic patient. The patient learned to open and close a prosthetic hand distant from his own hand with intention-driven neuronal ensemble activity. No improvements in voluntary motor function in the paralyzed hand were reported. Motor disability resulting from chronic stroke represents the main cause of long-term disability among adults and has substantial social, nancial, and psychological impact on patients, families, and society. Approximately one third of all stroke patients are not able to use the paralyzed hand for activities of daily living 1 year after the stroke. No treatment is available for that condition. A recent study (Buch et al., 2008) using a neuromagnetic BMI showed as a proof-of-principle successful BMI control of opening and closing grasping functions of an orthosis attached to the plegic hand in six out of eight patients. The orthosis was controlled by activity in three of the 275 MEG sensors. Increase of 912 Hz mu-rhythm in these three sensors opened the hand while decrease closed it. In six of the eight patients mu activity was derived from central ipsilesional location close to the subcortical lesion (see Fig. 4). After 1322 1-h training sessions, patients were able to control hand opening closing functions through the orthosis, in the absence of clinical improvements in the completely paralyzed hands. Training resulted in refocusing of MEG activity, providing rst evidence that BMI training may result in well-dened cortical reorganization. Whether an invasive BMI with implanted electrodes and internalized connection to the peripheral nerves, or noninvasive BMIs connected to prosthetic devices or rehabilitation robots may move from these bench type of study to the clinic awaits further research. Still, the gap between what can be achieved with implanted microelectrode arrays in motor or parietal cortex (Schwartz, 2006) in healthy nonhuman primates versus a paralyzed human patient is wide: While the monkey learns in relatively short time periods to use a small neural assembly to feed himself without any motor mediation, the human patient needs many training hours to open and close a paralyzed hand. The fact that a pattern of spiking neurons in the appropriate brain region is closer to the origin of movement production alone does not explain the explanatory gap: with a dense sensor array of MEG a complex four-directional hand movement was possible to reconstruct with an accuracy of 70% (Waldert et al., 2008) in healthy individuals. The prediction accuracy was only slightly smaller for EEG data.

Author's personal copy

114
A

BIRBAUMER et al.

PA

P
FIG. 4. Braincomputer interface (BCI) for motor restoration in chronic stroke. (A) MEG-BCI for control of mu-rhythm of an orthosis at the paralyzed hand, (B) mu-rhythm density (more yellowred) before (left) and after (right) 20 sessions of MEG-BCI-training. Black dots indicate MEG-sensors used for the BCI-training, (C) Percent correct responses (opening-closing hand) over 19 sessions of BCItraining in one patient, and (D) MR with subcortical region (right hemisphere) of the patient whose data are presented in (A)(C).

Experiments with lesioned animals and simultaneous recording of spike patterns, local eld potentials, and ECoG are urgently needed to explore the precise parameters at each level of observation necessary to reconstruct movements in the lesioned brain and/or the paralyzed body parts.

V. Conclusion

Despite a growing animal literature demonstrating online control of functional hand movements from spike patterns recorded with microelectrodes in the motor cortex, BMI applications in neurological patients are rare and hampered by methodological diYculties. BMIs using EEG-measures allow verbal communication in paralyzed patients with ALS, BMI-communication in totally locked-in patients, however, awaits experimental conrmation. Movement restoration in chronic stroke without residual movement capacity using noninvasive BMI is possible but generalization of improvement to real life needs further experimentation.

Author's personal copy

NEUROFEEDBACK AND BCI

115

Invasive and noninvasive BMIs using recordings from nerve cells, large neuronal pools such as ECoG and EEG, or blood ow-based measures such as f MRI and NIRS show potential for communication in LIS and movement restoration in chronic stroke, but controlled phase III clinical trials with larger populations of severely disturbed patients are urgently needed.

Acknowledgments

This work was supported by the Deutsche Forschungsgemeinschaft (DFG), Bundesministerium fur Bildung und Forschung (BMBF, Bernstein-Center for Neurotechnology 01GQ0831), Fatronik, San Sebastian, Spain, Motorike, Cesarea, Israel. Pedro Montoya was supported by Spanish Ministry of Science and European Funds (FEDER) (grant SEJ200762312).

References

Birbaumer, N. (2006). Breaking the silence: Brain-computer-interfaces (BCI) for communication and motor control. Psychophysiology 43, 517532. Birbaumer, N., and Cohen, L. (2007). Brain-computer-interfaces (BCI): Communication and restoration of movement in paralysis. J. Physiol. 579(3), 621636. Birbaumer, M., Elbert, T., Canavan, A., and Rockstroh, B. (1990). Slow potentials of the cerebral cortex and behavior. Physiol. Rev. 70, 141. Birbaumer, N., Ghanayim, N., Hinterberger, T., Iversen, I., Kotchoubey, B., Kubler, A., Perelmouter, J., Taub, E., and Flor, H. (1999). A spelling device for the paralysed. Nature 398, 297298. Buch, E., Weber, C., Cohen, L. G., Braun, C., Dimyan, M., Ard, T., Mellinger, J., Caria, A., Soekadar, S., Fourkas, A., and Birbaumer, N. (2008). Think to move: A neuromagnetic braincomputer interface (BCI) system for chronic stroke. Stroke 39, 910917. Caria, A., Veit, R., Sitaram, R., Lotze, M., Weiskopf, N., Grodd, W., and Birbaumer, N. (2007). Regulation of anterior insular cortex activity using real-time fMRI. NeuroImage 35, 12381246. DeCharms, R. C., Maeda, F., Glover, G. H., Ludlow, D., Pauly, J. M., Soneji, D., Gabrieli, J. D. E., and Mackey, S. C. (2005). Control over brain activation and pain learned by using real-time functional MRI. Proc. Natl. Acad. Sci. USA 102, 1862618631. Dornhege, C., Millan, J. R., Hinterberger, T., McFarland, D. J., and Muller, K. R. (Eds.) (2007). Toward Brain-Computer Interfacing. MIT Press, Cambridge, MA. Dworkin, B. R., and Miller, N. E. (1986). Failure to replicate visceral learning in the acute curarized rat preparation. Behav. Neurosci. 100, 299314. Elbert, T., Rockstroh, B., Lutzenberger, W., and Birbaumer, N. (Eds.) (1984). Self-Regulation of the Brain and Behavior. Springer, New York. Farwell, L. A., and Donchin, E. (1988). Talking oV the top of your head: Toward a mental prosthesis utilizing event-related brain potentials. Electroenc. Clin. Neurophysiol. 70, 510523.

Author's personal copy

116

BIRBAUMER et al.

Felton, E. A., Wilson, J. A., Williams, J. C., and Garell, P. C. (2007). Electrocorticographically controlled braincomputer interfaces using motor and sensory imagery in patients with temporary subdural electrode implants. J. Neurosurg. 106, 495500. Flor, H., and Birbaumer, N. (1993). Comparison of the eYcacy of EMG biofeedback, cognitive behavior therapy, and conservative medical interventions in the treatment of chronic musculoskeletal pain. J. Consult. Clin. Psychol. 61(4), 653658. Fuchs, T., Birbaumer, N., Lutzenberger, W., Gruzelier, J. H., and Kaiser, J. (2003). Neurofeedback training for attention-decit/hyperactivity disorder in children: A comparison with methylphenidate. Appl. Psychophysiol. Biofeedback 28(1), 112. Georgopoulos, A. P., Kalaska, J. F., Caminiti, R., and Massey, J. T. (2007). On the relations between the direction of two-dimensional arm movements and cell discharge in primate motor cortex. J. Neurophysiol. 2, 15271537. Hinterberger, T., Kubler, A., Kaiser, J., Neumann, N., and Birbaumer, N. (2003). A brain-computer interface (BCI) for the locked-in: Comparison of diVerent EEG classications for the thought translation device. Clin. Neurophysiol. 114, 416425. Hinterberger, T., Veit, R., Wilhelm, B., Weiskopf, N., Vatine, J. J., and Birbaumer, N. (2005). Neuronal mechanisms underlying control of a brain-computer interface. Eur. J. Neurosci. 21, 31693181. Hochberg, L. R., Serruya, M. D., Friehs, G. M., Mukand, J. A., Saleh, M., Caplan, A. H., Branner, A., Chen, D., Penn, R. D., and Donoghue, J. P. (2006). Neuronal ensemble control of prosthetic devices by a human with tetraplegia. Nature 442, 164171. Kotchoubey, B. (2005). Event-related potential measures of consciousness: Two equations with three unknowns. Prog. Brain Res. 150, 427444. Kotchoubey, B., Strehl, U., Uhlmann, C., Holzapfel, S., Konig, M., Froscher, W., Blankenhom, V., and Birbaumer, N. (2001). Modication of slow cortical potentials in patients with refractory epilepsy: A controlled outcome study. Epilepsia 42(3), 406416. Kubler, A., and Birbaumer, N. (2008). Brain-computer interfaces and communication in paralysis: Extinction of goal directed thinking in completely paralysed patients? Clin. Neurophysiol. 119(11), 26582666. Lal, T. N., Schroder, M., Hinterberger, T., Weston, J., Bogdan, M., Birbaumer, N., and Scholkopf, B. (2004). Support vector channel selection in BCI. IEEE Trans. Biomed. Eng. 51(6), 10031010. Lebedev, M. A., and Nicolelis, M. A. (2006). Brain machine interfaces: Past, present and future. Trends Neurosci. 29(9), 536546. Miller, N. E. (1969). Learning of visceral and glandular responses. Science 163, 434445. Nagai, Y., Goldstein, L. H., Fenwick, P. B. C., and Trimble, M. R. (2004). Clinical eYcacy of galvanic skin response biofeedback training in reducing seizures in adult epilepsy: A preliminary randomized controlled study. Epilepsy Behav. 5, 216223. Naito, M., Michioka, Y., Ozawa, K., et al. (2007). A communication means for totally locked-in ALS patients based on changes in cerebral blood volume measured with near-infrared light. IEICE Trans. Inf. Syst. E90-D, 10281037. Nicolelis, M. A. L., Birbaumer, N., and Mueller, K.-L. (2004). Brain-computer interfaces. Guest Editorial. IEEE Trans. Biomed. Eng. 51, 877880. Nijboer, F., Furdea, A., Gunst, I., Mellinger, J., McFarland, D. J., Birbaumer, N., and Kubler, A. (2008). An auditory brain-computer-interface (BCI). J. Neurosci. Methods 167, 4350. Nijboer, F., Sellers, E. W., Mellinger, J., Jordan, M. A., Matuz, T., Halder, S., Mochty, U., Krusienski, D. J., Vaughan, T. M., Wolpaw, J. R., Birbaumer, N., Kubler, A. (2008). A P300 based brain-computer interface for people with amyotrophic lateral sclerosis. Clin. Neurophysiol. 119(8), 19091916. Owen, A. M., Coleman, M. R., Boly, M., Melanie, B., Matthew, H. D., Steven, L., and John, D. P. (2006). Detecting awareness in the vegetative state. Science 313(5792), 1402.

Author's personal copy

NEUROFEEDBACK AND BCI

117

Pfurtscheller, G., Guger, C., Muller, G., Krausz, G., and Neuper, C. (2000). Brain oscillations control hand orthosis in a tetraplegic. Neurosci. Lett. 292(3), 211214. Pfurtscheller, G., Neuper, C., and Birbaumer, N. (2005). Human brain-computer interface (BCI). In Motor Cortex in Voluntary Movements. A Distributed System for Distributed Functions (A. Riehle and E. Vaadia, Eds.), pp. 367401. CRC Press, Boca Raton. Riehle, A., and Vaadia, E. (Eds.) (2005). Motor Cortex in Voluntary Movements. A Distributed System for Distributed Functions. CRC Press, Boca Raton. Rockstroh, B., Elbert, T., Birbaumer, N., Wolf, P., Duchting-Roth, A., Reker, M., Daum, I., Lutzenberger, W., and Dichgans, J. (1993). Cortical self-regulation in patients with epilepsies. Epilepsy Res. 14, 6372. Schwartz, A. B. (2006). Neurobiology: Crossed circuits. Nature 444(7115), 4748. Schwartz, M., and Andrasik, F. (Eds.) (2003). Biofeedback: A Practitioners Guide. 3rd ed. Guilford Press, New York. Sellers, E. W., and Donchin, E. (2006). A P300-based brain-computer interface: Initial tests by ALS patients. Clin. Neurophysiol. 117, 538548. Sitaram, R., Zhang, H., Guan, C., Thulasidas, M., Hoshi, Y., Ishikawa, A., Shimizu, K., and Birbaumer, N. (2007). Temporal classication of multi-channel near-infrared spectroscopy signals of motor imagery for developing a brain-computer interface. NeuroImage 34, 14161427. Sterman, M. B., and Clemente, C. D. (1962a). Forebrain inhibitory mechanisms: Cortical synchronization induced by basal forebrain stimulation. Exp. Neurol. 6, 91102. Strehl, U., Leins, U., Goth, G., Klinger, C., Hinterberger, T., and Birbaumer, N. (2006). Self-regulation of slow cortical potentialsA new treatment for children with attention-decit/hyperactivity disorder. Pediatrics 118(5), 15301540. Velliste, M., Perel, S., Spalding, M. C., et al. (2008). Cortical control of a prosthetic arm for self-feeding. Nature 453, 10981101, doi:10.1038/nature06996. Waldert, S., Preissl, H., Demandt, E., Braun, C., Birbaumer, N., Aertsen, A., and Mehring, C. (2008). Hand movement direction decoded from MEG and EEG. J. Neurosci. 28(4), 10001008. Walker, J. E., and Kozlowski, G. (2005). Neurofeedback training of epilepsy. Child Adolesc. Psychiatry Clin. N. Am. 14, 163176. Weiskopf, N., Scharnowski, F., Veit, R., Goebel, R., Birbaumer, N., and Mathiak, K. (2005). Self-regulation of local brain activity using real-time functional magnetic resonance imaging (f MRI). J. Physiol. Paris 98, 357373.