Modeling and analysis of mechanisms underlying fMRI-based decoding of information conveyed in cortical columns

Abstract

Multivariate machine learning algorithms applied to human functional MRI (fMRI) data can decode information conveyed by cortical columns, despite the voxel-size being large relative to the width of columns. Several mechanisms have been proposed to underlie decoding of stimulus orientation or the stimulated eye. These include: (I) aliasing of high spatial-frequency components, including the main frequency component of the columnar organization, (II) contributions from local irregularities in the columnar organization, (III) contributions from large-scale non-columnar organizations, (IV) functionally selective veins with biased draining regions, and (V) complex spatio-temporal filtering of neuronal activity by fMRI voxels. Here we sought to assess the plausibility of two of the suggested mechanisms: (I) aliasing and (II) local irregularities, using a naive model of BOLD as blurring and MRI voxel sampling. To this end, we formulated a mathematical model that encompasses both the processes of imaging ocular dominance (OD) columns and the subsequent linear classification analysis. Through numerical simulations of the model, we evaluated the distribution of functional differential contrasts that can be expected when considering the pattern of cortical columns, the hemodynamic point spread function, the voxel size, and the noise. We found that with data acquisition parameters used at 3 Tesla, sub-voxel supra-Nyquist frequencies, including frequencies near the main frequency of the OD organization (0.5 cycles per mm), cannot contribute to the differential contrast. The differential functional contrast of local origin is dominated by low-amplitude contributions from low frequencies, associated with irregularities of the cortical pattern. Realizations of the model with parameters that reflected best-case scenario and the reported BOLD point-spread at 3 Tesla (3.5 mm) predicted decoding performances lower than those that have been previously obtained at this magnetic field strength. We conclude that low frequency components that underlie local irregularities in the columnar organization are likely to play a role in decoding. We further expect that fMRI-based decoding relies, in part, on signal contributions from large-scale, non-columnar functional organizations, and from complex spatio-temporal filtering of neuronal activity by fMRI voxels, involving biased venous responses. Our model can potentially be used for evaluating and optimizing data-acquisition parameters for decoding information conveyed by cortical columns.

Separate neural representations for physical pain and social rejection

“…Among methods available in healthy humans, fMRI multivariate patterns may be particularly useful in inferring representational similarity 43 , as such patterns can be sensitive to population codes distributed across large numbers of individual neurons. FMRI multivariate patterns have obvious limitations in spatial resolution compared with single-unit recording 49 , in large part because blood-oxygen-level-dependent (BOLD) fMRI is sensitive to changes in microvascular beds that subserve many neurons, and the BOLD response thus acts as a spatial low-pass (blurring) filter sensitive to some, but not all, neuronal-level processes 48,50 . Because of the spatial blurring properties of fMRI, finding that two manipulations activate similar fMRI patterns does not necessarily imply similar neuron-level representations.…”

Implicit representations of luminance and the temporal structure of moving stimuli in multiple regions of human visual cortex revealed by multivariate pattern classification analysis

“…Together, this suggests that the effect of eccentricity on speed preference did not bias classifier performance, even if the classifier may, in part, have relied on information contained in these large-scale biases. Although there is controversy about the spatial scale of BOLD responses underlying MVPA (Freeman et al 2011; Kamitani and Sawahat 2010), it is likely that information in the BOLD response may be found at several spatial scales, all of which may drive classification performance (Chaimow et al 2011). The fact that the speed classification analysis did not show a sensitivity to luminance could thus reflect the existence of smaller-scale speed biases in the multivariate response, which might have been invariant to luminance differences.…”

What the success of brain imaging implies about the neural code

“…This result is in line with basic concepts from information theory, such as the Nyquist-Shannon sampling theorem. The key is that the red and yellow pixels/neurons are topologically organized: their relationship to each other is for all intents and purposes invariant to the granularity of the squares/voxels (for more details see: Chaimow et al, 2011; Freeman et al, 2011; Swisher et al, 2010).

10.7554/eLife.21397.003Figure 1.The activity of neurons in the top-left panel gradually changes from left to right, whereas changes are more abrupt in the top-middle and top-right panels.Each square in the grid represents a voxel which summates activity within its frame as shown in the bottom panels.

…”