Search

At ultra-high field, fMRI voxels can span the sub-millimeter range, allowing the recording of blood oxygenation level dependent (BOLD) responses at the level of fundamental units of neural computation, such as cortical columns and layers. This sub-millimeter resolution, however, is only nominal in nature as a number of factors limit the spatial acuity of functional voxels. Multivoxel Pattern Analysis (MVPA) may provide a means to detect information at finer spatial scales that may otherwise not be visible at the single voxel level due to limitations in sensitivity and specificity. Here, we evaluate the spatial scale of stimuli specific BOLD responses in multivoxel patterns exploited by linear Support Vector Machine, Linear Discriminant Analysis and Naïve Bayesian classifiers across cortical depths in V1. To this end, we artificially misaligned the testing relative to the training portion of the data in increasing spatial steps, then investigated the breakdown of the classifiers' performances. A one voxel shift led to a significant decrease in decoding accuracy (p < 0.05) across all cortical depths, indicating that stimulus specific responses in a multivoxel pattern of BOLD activity exploited by multivariate decoders can be as precise as the nominal resolution of single voxels (here 0.8 mm isotropic). Our results further indicate that large draining vessels, prominently residing in proximity of the pial surface, do not, in this case, hinder the ability of MVPA to exploit fine scale patterns of BOLD signals. We argue that tailored analytical approaches can help overcoming limitations in high-resolution fMRI and permit studying the mesoscale organization of the human brain with higher sensitivities.

Clark offers a powerful description of the brain as a prediction machine, which offers progress on two distinct levels. First, on an abstract conceptual level, it provides a unifying framework for perception, action, and cognition (including subdivisions such as attention, expectation, and imagination). Second, hierarchical prediction offers progress on a concrete descriptive level for testing and constraining conceptual elements and mechanisms of predictive coding models (estimation of predictions, prediction errors, and internal models).

Competent social organisms will read the social signals of their peers. In primates, the face has evolved to transmit the organism's internal emotional state. Adaptive action suggests that the brain of the receiver has co-evolved to efficiently decode expression signals. Here, we review and integrate the evidence for this hypothesis. With a computational approach, we co-examined facial expressions as signals for data transmission and the brain as receiver and decoder of these signals. First, we show in a model observer that facial expressions form a lowly correlated signal set. Second, using time-resolved EEG data, we show how the brain uses spatial frequency information impinging on the retina to decorrelate expression categories. Between 140 to 200 ms following stimulus onset, independently in the left and right hemispheres, an information processing mechanism starts locally with encoding the eye, irrespective of expression, followed by a zooming out to processing the entire face, followed by a zooming back in to diagnostic features (e.g. the opened eyes in "fear", the mouth in "happy"). A model categorizer demonstrates that at 200 ms, the left and right brain have represented enough information to predict behavioral categorization performance.

A key to understanding visual cognition is to determine when, how, and with what information the human brain distinguishes between visual categories. So far, the dynamics of information processing for categorization of visual stimuli has not been elucidated. By using an ecologically important categorization task (seven expressions of emotion), we demonstrate, in three human observers, that an early brain event (the N170 Event Related Potential, occurring 170 ms after stimulus onset) integrates visual information specific to each expression, according to a pattern. Specifically, starting 50 ms prior to the ERP peak, facial information tends to be integrated from the eyes downward in the face. This integration stops, and the ERP peaks, when the information diagnostic for judging a particular expression has been integrated (e.g., the eyes in fear, the corners of the nose in disgust, or the mouth in happiness). Consequently, the duration of information integration from the eyes down determines the latency of the N170 for each expression (e.g., with "fear" being faster than "disgust," itself faster than "happy"). For the first time in visual categorization, we relate the dynamics of an important brain event to the dynamics of a precise information-processing function.