Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

Co-Planar Stereotaxic Atlas of the Human Brain—3-Dimensional Proportional System: An Approach to Cerebral Imaging, J. Talairach, P. Tournoux. Georg Thieme Verlag, New York (1988), 122 pp., 130 figs. DM 268

I have developed "tennis elbow" from lugging this book around the past four weeks, but it is worth the pain, the effort, and the aspirin. It is also worth the (relatively speaking) bargain price. Including appendixes, this book contains 894 pages of text. The entire panorama of the neural sciences is surveyed and examined, and it is comprehensive in its scope, from genomes to social behaviors. The editors explicitly state that the book is designed as "an introductory text for students of biology, behavior, and medicine," but it is hard to imagine any audience, interested in any fragment of neuroscience at any level of sophistication, that would not enjoy this book. The editors have done a masterful job of weaving together the biologic, the behavioral, and the clinical sciences into a single tapestry in which everyone from the molecular biologist to the practicing psychiatrist can find and appreciate his or

Principles of Neural Science

Neuromotor prostheses (NMPs) aim to replace or restore lost motor functions in paralysed humans by routeing movement-related signals from the brain, around damaged parts of the nervous system, to external effectors. To translate preclinical results from intact animals to a clinically useful NMP, movement signals must persist in cortex after spinal cord injury and be engaged by movement intent when sensory inputs and limb movement are long absent. Furthermore, NMPs would require that intention-driven neuronal activity be converted into a control signal that enables useful tasks. Here we show initial results for a tetraplegic human (MN) using a pilot NMP. Neuronal ensemble activity recorded through a 96-microelectrode array implanted in primary motor cortex demonstrated that intended hand motion modulates cortical spiking patterns three years after spinal cord injury. Decoders were created, providing a ‘neural cursor’ with which MN opened simulated e-mail and operated devices such as a television, even while conversing. Furthermore, MN used neural control to open and close a prosthetic hand, and perform rudimentary actions with a multi-jointed robotic arm. These early results suggest that NMPs based upon intracortical neuronal ensemble spiking activity could provide a valuable new neurotechnology to restore independence for humans with paralysis. The cover shows Matt Nagle, first participant in the BrainGate pilot clinical trial. He is unable to move his arms or legs following cervical spinal cord injury. Researchers at the Department of Neuroscience at Brown University, working with biotech company Cyberkinetics and 3 other institutions, demonstrate that movement-related signals can be relayed from the brain via an implanted BrainGate chip, allowing the patient to drive a computer screen cursor and activate simple robotic devices. Such neuromotor prostheses could pave the way towards systems to replace or restore lost motor function in paralysed humans. Prior to this advance, this type of work has been performed chiefly in monkeys. The latest example of such research has achieved a large increase in speed over current devices, enhancing the prospects for clinically viable brain-machine interfaces.

Neuronal ensemble control of prosthetic devices by a human with tetraplegia

Over the past 20 years, neuroimaging has become a predominant technique in systems neuroscience. One might envisage that over the next 20 years the neuroimaging of distributed processing and connectivity will play a major role in disclosing the brain's functional architecture and operational principles. The inception of this journal has been foreshadowed by an ever-increasing number of publications on functional connectivity, causal modeling, connectomics, and multivariate analyses of distributed patterns of brain responses. I accepted the invitation to write this review with great pleasure and hope to celebrate and critique the achievements to date, while addressing the challenges ahead.

/pdf/functional-and-effective-connectivity-a-review-s5oik3vige.pdf

Functional and effective connectivity: a review.

We investigated how the assumed hand position of the right arm during movement imagination correlates with brain sources. Sources were calculated from the electroencephalogram (EEG) applying a brain imaging technique. We used frequencies below 1 Hz. In 4 out of 9 subjects, substantial correlations on the supplementary motor cortex were observed. This supports that decoding movement imaginations could be possible and it is neurophysiologically plausible.

Time-Domain Correlations of Imagined Arm Positions with Brain Sources.

Introduction: Operating brain-computer interfaces (BCIs) that are based on modulation of spontaneous electroencephalogram (EEG) rhythms is a skill that users have to learn. Training time varies significantly among users and it can take up to several months of time before minimum BCI control becomes possible. One reason for this may be the design of BCI training paradigms [1]. Trials of mental activity are collected prior to BCI use and used to calibrate BCI model parameters. During this initial data collection no feedback is provided to users. This is not very motivating and engaging for users. Recently, we introduced a thought-based one-switch row-column communication board for users with cerebral palsy [2]. To make BCI skill learning more motivating, we lately also introduced a jigsaw puzzle game-based training paradigm [3] that presents feedback from the very start. Initially sham feedback and once enough data is available for BCI calibration, online feedback. Moreover, the game was designed following user-centered design principles and explains the concept of BCI use and allows users to get familiar with the dynamics of BCI control. A pilot study in four users with cerebral palsy (CP) showed that the acceptance was high and users liked the game-based training approach [3]. Three out of the four users achieved performance better than chance level after about 30 minutes of training. Fig. 1 shows a user assembling a puzzle. Here, we propose social inclusion as another feature for improving BCI training. End users should not perform training alone but engage in direct competition with other persons.

Social inclusion as feature to improve BCI skill training: A feasibility case study in cerebral palsy

Introduction: Operating a brain-computer interface (BCI) that decodes spontaneous electroencephalogram (EEG) rhythms is a skill that users have to learn [1,2]. Feedback is essential for skill learning. EEG is typically recorded from users prior to BCI use and used to train pattern recognition models (calibration). During online operation, the model output is commonly presented as feedback. We aim at developing BCIs that do not require calibration. To this end we study between-user model transfer and online co-adaptation. The former allows providing feedback from the start and the latter enables user-specific finetuning of model parameters to enhance performance. The aim of this offline study is to evaluate the usefulness of the Restricted Boltzmann Machine (RBM) model as a generative classification framework [3] for the above mentioned purposes.

Skipping BCI calibration: fundamental investigations on Restricted Boltzmann Machines

In the present study, we aimed to detect the ”resonance-like” frequencies of the somatosensory system in patients in a minimally conscious state using a screening paradigm. EEG measurements were conducted in seven patients during tactile stimulation of their left and right wrist. A significant tuning curve could be found in one of the patients. Various reasons that could explain the inconclusive outcome of most measurements, as well as future perspectives are discussed.

Gernot Müller-Putz

Papers

Time-Domain Correlations of Imagined Arm Positions with Brain Sources.

Social inclusion as feature to improve BCI skill training: A feasibility case study in cerebral palsy

Skipping BCI calibration: fundamental investigations on Restricted Boltzmann Machines

Steady-State Somatosensory Evoked Potentials in Minimally Conscious Patients – Challenges and Perspectives

Combination of Manual Control and Motor Imagery to Move an Artificial Limb