Joaquin Rapela

Internet2-based 3D PET image reconstruction using a PC cluster

We describe an approach to fast iterative reconstruction from fully three-dimensional (3D) PET data using a network of PentiumIII PCs configured as a Beowulf cluster. To facilitate the use of this system, we have developed a browser-based interface using Java. The system compresses PET data on the users machine, sends these data over a network, and instructs the PC cluster to reconstruct the image. The cluster implements a parallelized version of our preconditioned conjugate gradient method for fully 3D MAP image reconstruction. We report on the speed-up factors using the Beowulf approach and the impacts of communication latencies in the local cluster network and the network connection between the users machine and our PC cluster.

Estimating nonlinear receptive fields from natural images.

The response of visual cells is a nonlinear function of their stimuli. In addition, an increasing amount of evidence shows that visual cells are optimized to process natural images. Hence, finding good nonlinear models to characterize visual cells using natural stimuli is important. The Volterra model is an appealing nonlinear model for visual cells. However, their large number of parameters and the limited size of physiological recordings have hindered its application. Recently, a substantiated hypothesis stating that the responses of each visual cell could depend on an especially low-dimensional subspace of the image space has been proposed. We use this low-dimensional subspace in the Volterra relevant-space technique to allow the estimation of high-order Volterra models. Most laboratories characterize the response of visual cells as a nonlinear function on the low-dimensional subspace. They estimate this nonlinear function using histograms and by fitting parametric functions to them. Here, we compare the Volterra model with these histogram-based techniques. We use simulated data from cortical simple cells as well as simulated and physiological data from cortical complex cells. Volterra models yield equal or superior predictive power in all conditions studied. Several methods have been proposed to estimate the low-dimensional subspace. In this article, we test projection pursuit regression (PPR), a nonlinear regression algorithm. We compare PPR with two popular models used in vision: spike-triggered average (STA) and spike-triggered covariance (STC). We observe that PPR has advantages over these alternative algorithms. Hence, we conclude that PPR is a viable algorithm to recover the relevant subspace from natural images and that the Volterra model, estimated through the Volterra relevant-space technique, is a compelling alternative to histogram-based techniques.

ePPR: a new strategy for the characterization of sensory cells from input/output data

A central goal of systems neuroscience is to characterize the transformation of sensory input to spiking output in single neurons. This problem is complicated by the large dimensionality of the inputs. To cope with this problem, previous methods have estimated simplified versions of a generic linear-nonlinear (LN) model and required, in most cases, stimuli with constrained statistics. Here we develop the extended Projection Pursuit Regression (ePPR) algorithm that allows the estimation of all of the parameters, in space and time, of a generic LN model using arbitrary stimuli. We first prove that ePPR models can uniformly approximate, to an arbitrary degree of precision, any continuous function. To test this generality empirically, we use ePPR to recover the parameters of models of cortical cells that cannot be represented exactly with an ePPR model. Next we evaluate ePPR with physiological data from primary visual cortex, and show that it can characterize both simple and complex cells, from their responses to both natural and random stimuli. For both simulated and physiological data, we show that ePPR compares favorably to spike-triggered and information-theoretic techniques. To the best of our knowledge, this article contains the first demonstration of a method that allows the estimation of an LN model of visual cells, containing multiple spatio-temporal filters, from their responses to natural stimuli.

Automatically combining ranking heuristics for HTML documents

Current search engines use several criteria or heuristics to rank HTML documents. HTML ranking heuristics need to be combined into a ranking function that given a text query returns a ranked list of HTML documents. The standard approach is to build a weighted average by manually estimating the importance of every heuristic and assigning a weight proportional to the estimated importance. In the current paper we apply an automatic method for combining HTML ranking heuristics. Using recall/precision evaluations we study the performance of the automatic method and using collections of HTML documents with different characteristics we show that the automatic method finds weights tailored to specific characteristics of each document collection

Assisting autistic children with wireless EOG technology

We propose a novel intervention to train the speed and accuracy of attention orienting and eye movements in Autism Spectrum Disorder (ASD). Training eye movements and attention could not only affect those important functions directly, but could also result in broader improvement of social communication skills. To this end we describe a system that would allow ASD children to improve their fixation skills while playing a computer game controlled by an eye tracker. Because this intervention will probably be time consuming, this system should be designed to be used at homes. To make this possible, we propose an implementation based on wireless and dry electrooculography (EOG) technology. If successful, this system would develop an approach to therapy that would improve clinical and behavioral function in children and adults with ASD. As our initial steps in this direction, here we describe the design of a computer game to be used in this system, and the predictions of gaze position from EOG data recorded while a subject played this game.

Brain oscillations in switching vs. focusing audio-visual attention

Selective attention contributes to perceptual efficiency by modulating cortical activity according to task demands. The majority of attentional research has focused on the effects of attention to a single modality, and little is known about the role of attention in multimodal sensory processing. Here we employ a novel experimental design to examine the electrophysiological basis of audio-visual attention shifting. We use electroencephalography (EEG) to study differences in brain dynamics between quickly shifting attention between modalities and focusing attention on a single modality for extended periods of time. We also address interactions between attentional effects generated by the attention-shifting cue and those generated by subsequent stimuli. The conclusions from these examinations address key issues in attentional research, including the supramodal theory of attention, or the role of attention in foveal vision. The experimental design and analysis methods used here may suggest new directions in the study of the physiological basis of attention.

A new foreperiod effect on single-trial phase coherence. Part I: existence and relevance

Expecting events in time leads to more efficient behavior. A remarkable early finding in the study of temporal expectancy is the foreperiod effect on reaction times; i.e., the fact that the time period between a warning signal and an impendent stimuli, to which subjects are instructed to respond as quickly as possible, influences reaction times. Recently it has been shown that the phase of oscillatory activity preceding stimulus presentation is related to behavior. Here we connect both of these findings by reporting a novel foreperiod effect on the inter-trial phase coherence triggered by a stimulus to which subjects do not respond. Until now, inter-trial phase coherence has been used to describe a regularity in the phases of groups of trials. We propose a single-trial measure of inter-trial phase coherence and prove its soundness. Equipped with this measure, and using a multivariate decoding method, we demonstrate that the foreperiod duration modulates single-trial phase coherence. In principle, this modulation could be an artifact due to the decoding method used to detect it. We show that this is not the case, since the modulation can also be observed with a very simple averaging method. Although real, the single-trial modulation of inter-trial phase coherence by the foreperiod duration could just reflect a nuisance in our data. We argue against this possibility by showing that the strength of the modulation correlates with subjects’ behavioral measures, both error rates and mean-reaction times. We anticipate that the new foreperiod effect on inter-trial phase coherence, and the decoding method used here to detect it, will be important tools to understand cognition at the single-trial level. In Part II of this manuscript, we support this claim, by showing that attention modulates the strength of the new foreperiod effect in a trial-by-trial basis.