Optimal unsupervised motor learning for dimensionality reduction of nonlinear control systems

In this paper, optimal unsupervised motor learning is defined to be a technique for finding the coordinate system of minimum dimensionality which can adequately describe a particular motor task. An explicit method is provided for learning a stable controller that translates commands within the new coordinate system into motor variables appropriate for plant control. The method makes use of previously described neural network algorithms including the generalized Hebbian algorithm, basis-function trees, and trajectory extension learning. Examples of applications to a real direct-drive two joint planar robot arm and a simulated three joint robot arm with visual sensing are given.

Expanding CEP290 mutational spectrum in ciliopathies

Ciliopathies are an expanding group of rare conditions characterized by multiorgan involvement, that are caused by mutations in genes encoding for proteins of the primary cilium or its apparatus. Among these genes, CEP290 bears an intriguing allelic spectrum, being commonly mutated in Joubert syndrome and related disorders (JSRD), Meckel syndrome (MKS), Senior-Loken syndrome and isolated Leber congenital amaurosis (LCA). Although these conditions are recessively inherited, in a subset of patients only one CEP290 mutation could be detected. To assess whether genomic rearrangements involving the CEP290 gene could represent a possible mutational mechanism in these cases, exon dosage analysis on genomic DNA was performed in two groups of CEP290 heterozygous patients, including five JSRD/MKS cases and four LCA, respectively. In one JSRD patient, we identified a large heterozygous deletion encompassing CEP290 C-terminus that resulted in marked reduction of mRNA expression. No copy number alterations were identified in the remaining probands. The present work expands the CEP290 genotypic spectrum to include multiexon deletions. Although this mechanism does not appear to be frequent, screening for genomic rearrangements should be considered in patients in whom a single CEP290 mutated allele was identified.

A writing device improves writing in primary writing tremor

Primary writing tremor (PWT) is task-specific and interferes with handwriting. Several reports have shown a beneficial response of this disorder to stereotactic functional neurosurgery. Significant improvement with a writing device was demonstrated with blind rating of handwriting and spiral drawing samples collected before and during its use in nine patients with PWT, suggesting that this therapeutic modality should be tried before considering chronic pharmacotherapy or functional neurosurgery.

Childhood onset generalised dystonia can be modelled by increased gain in the indirect basal ganglia pathway

Clinical experience suggests an important role of the indirect basal ganglia pathway in the genesis of childhood onset generalised dystonia, but it has been difficult to reconcile the increased muscle activity in dystonia with the current model of basal ganglia function in which the indirect pathway is considered primarily inhibitory. The aim of this study was to present a modification of the direct-indirect pathway model, in which the indirect pathway is inverting rather than purely inhibitory, so that while high signals are inhibited, low signals are amplified. As the basal ganglia may be a feedback loop that modifies cortical activity, instability from excessive gain in this feedback loop could explain features of dystonia. A detailed mathematical model is provided, together with simulations of cortical cell population spiking behaviour when connected through a basal ganglia loop. The simulations show that increased gain in the indirect pathway relative to the direct pathway can lead to unstable uncontrolled synchronous oscillations in cortex and basal ganglia. This behaviour could result in dystonia. The model provides a consistent explanation for the association of dystonia with parkinsonism and disorders characterised by dopamine depletion, the ability to treat some dystonias with dopamine, the ability of neuroleptic drug treatment to cause an acute dystonic reaction treatable with anticholinergic drugs, and the ability of pallidotomy or deep brain stimulation of the internal pallidum to alleviate symptoms of generalised dystonia.

Interactions between two different inhibitory systems in the human motor cortex

Intracortical inhibition in the human motor cortex has been previously demonstrated using paired-pulse transcranial magnetic stimulation (TMS) protocols at short intervals (1-6 ms; short interval intracortical inhibition, SICI) with a subthreshold conditioning pulse preceding a suprathreshold test pulse, and at long intervals (50-200 ms; long interval intracortical inhibition, LICI) with suprathreshold conditioning and test pulses. We investigated whether different circuits mediate these inhibitory phenomena and how they interact. In nine healthy volunteers, we applied TMS to the motor cortex and recorded motor evoked potentials from the first dorsal interosseous muscle. With increasing test pulse strength, LICI decreases but SICI tends to increase. There was no correlation between the degree of SICI and LICI. We tested the interactions between SICI and LICI. SICI was reduced or eliminated in the presence of LICI. Loss of SICI was seen even with a conditioning stimulus too weak to induce significant LICI. Our findings demonstrate that different cell populations mediate SICI and LICI. The results are consistent with the hypothesis that LICI inhibits SICI through presynaptic GABAB receptors. Testing of SICI in the presence of LICI may be a non-invasive way of evaluating inhibitory interactions in the human motor cortex.

Abnormalities of spatial and temporal sensory discrimination in writer's cramp

Clinical observations of patients with writer's cramp suggest that abnormalities of the sensory system may be a frequent finding in this disorder. Neurophysiological data from an animal model of focal dystonia have revealed cells in somatosensory cortex with enlarged and overlapping tactile receptive fields. However, psychophysical studies so far have been unable to document a clinical correlate supporting a similar enlargement of receptive fields in humans. We compared the fingertip discrimination of the orientation of fine spatial gratings between writer's cramp and control subjects and found a significant decrease in grating sensitivity in the patients, consistent with the possibility of enlarged tactile receptive fields. In addition, we duplicated previous experiments showing an abnormality of tactile temporal discrimination. The results provide psychophysical measures which may relate to the development of sensory cortical reorganization in patients with writer's cramp.

Computational model of the role of sensory disorganization in focal task-specific dystonia

We present a new computational model for the development of task-specific focal dystonia. The purpose of the model is to explain how altered sensory representations can lead to abnormal motor behavior. Dystonia is described as the result of excessive gain through a sensorimotor loop. The gain is determined in part by the sensory cortical area devoted to each motor function, and behaviors that lead to abnormal increases in sensory cortical area are predicted to lead to dystonia. Properties of dystonia including muscular co-contraction, overflow movements, and task specificity are predicted by properties of a linear approximation to the loop transformation. We provide simulations of several different mechanisms that can cause the gain to exceed 1 and the motor activity to become sustained and uncontrolled. The model predicts that normal plasticity mechanisms may contribute to worsening of symptoms over time.

Sensory discrimination capabilities in patients with focal hand dystonia

To explore the concept that dystonia may result from dysfunction of the sensory system, 14 patients with focal hand dystonia were tested during two somatosensory discrimination tasks. Compared with controls, patients had a higher threshold in a task involving discrimination of two electric stimuli closely related temporally, an abnormality that correlated with the degree of severity of dystonia. There was no significant difference in a single-touch, gross localization task. The possible relevance of these findings to the pathogenesis of dystonia is discussed.

Human arm movements described by a low-dimensional superposition of principal components

A new method for analyzing kinematic patterns during smooth movements is proposed. Subjects are asked to move the end of a two-joint manipulandum to copy a smooth initial target path. On subsequent trials the target path is the subject's actual movement from the preceding trial. Using Principal Components Analysis, it is shown that the trajectories have very low dimension and that they converge toward a linear superposition of the first few principal components. We show similar results for handwriting on an electronic pen tablet. We hypothesize that the low dimensionality and convergence are attributable to combined properties of the internal controller and the musculoskeletal system. The low dimensionality may allow for efficient descriptions of a large class of arm movements.

Probability density methods for smooth function approximation and learning in populations of tuned spiking neurons

This article proposes a new method for interpreting computations performed by populations of spiking neurons. Neural firing is modeled as a rate-modulated random process for which the behavior of a neuron in response to external input can be completely described by its tuning function. I show that under certain conditions, cells with any desired tuning functions can be approximated using only spike coincidence detectors and linear operations on the spike output of existing cells. I show examples of adaptive algorithms based on only spike data that cause the underlying cell-tuning curves to converge according to standard supervised and unsupervised learning algorithms. Unsupervised learning based on principal components analysis leads to independent cell spike trains. These results suggest a duality relationship between the random discrete behavior of spiking cells and the deterministic smooth behavior of their tuning functions. Classical neural network approximation methods and learning algorithms based on continuous variables can thus be implemented within networks of spiking neurons without the need to make numerical estimates of the intermediate cell firing rates.

Probability density estimation for the interpretation of neural population codes

1. Electrophysiological recording data from multiple cells in motor cortex and elsewhere often are interpreted using the population vector method pioneered by Georgopoulos and coworkers. This paper proposes an alternative method for interpreting coding across populations of cells that may succeed under circumstances in which the population vector fails. 2. Population codes are analyzed using probability theory to find the complete conditional probability density of a movement parameter given the firing pattern of a set of cells. 3. The conditional probability density when a single cell fires is proportional to the shape of the cell's tuning curve of firing rate in response to different movement parameters. 4. The conditional density when multiple cells fire is proportional to the product of their tuning curves. 5. Movement parameters can be estimated from the conditional density using statistical maximum likelihood or minimum mean-squared error methods. 6. Simulations show that density estimation correctly finds movement directions for nonuniform distributions of preferred directions and noncosine cell tuning curves, whereas the population vector method fails for these cases. 7. Probability methods thus provide a statistically based alternative to the population vector for interpreting electrophysiological recording data from multiple cells.

MERRF syndrome with overwhelming lactic acidosis

Myoclonic epilepsy with ragged-red fiber syndrome has been associated with a mitochondrial DNA base substitution at nucleotide 8344 in the mitochondrial tRNA(Lys) gene. In several reported series, adult patients with these mutations have mild to moderate symptoms that progress slowly over many years. We describe a girl with MERRF syndrome who had an unusually rapid and severe clinical course, with onset of symptoms at age 7 years and death by age 14 years of overwhelming lactic acidosis. Postmortem tissue biopsy revealed variable but generally high percentages of mutant mitochondrial genomes in multiple organ systems. Twenty-one other members of her family were tested for the mutation and had varying percentages in leukocytes.

Equilibrium point control of a monkey arm simulator by a fast learning tree structured artificial neural network

A planar 17 muscle model of the monkey's arm based on realistic biomechanical measurements was simulated on a Symbolics Lisp Machine. The simulator implements the equilibrium point hypothesis for the control of arm movements. Given initial and final desired positions, it generates a minimum-jerk desired trajectory of the hand and uses the backdriving algorithm to determine an appropriate sequence of motor commands to the muscles (Flash 1987; Mussa-Ivaldi et al. 1991; Dornay 1991b). These motor commands specify a temporal sequence of stable (attractive) equilibrium positions which lead to the desired hand movement. A strong disadvantage of the simulator is that it has no memory of previous computations. Determining the desired trajectory using the minimum-jerk model is instantaneous, but the laborious backdriving algorithm is slow, and can take up to one hour for some trajectories. The complexity of the required computations makes it a poor model for biological motor control. We propose a computationally simpler and more biologically plausible method for control which achieves the benefits of the backdriving algorithm. A fast learning, tree-structured network (Sanger 1991c) was trained to remember the knowledge obtained by the backdriving algorithm. The neural network learned the nonlinear mapping from a 2-dimensional cartesian planar hand position (x,y) to a 17-dimensional motor command space (u1, . . ., u17). Learning 20 training trajectories, each composed of 26 sample points [[x,y], [u1, . . ., u17] took only 20 min on a Sun-4 Sparc workstation. After the learning stage, new, untrained test trajectories as well as the original trajectories of the hand were given to the neural network as input. The network calculated the required motor commands for these movements. The resulting movements were close to the desired ones for both the training and test cases.