Multivoxel pattern analysis (MVPA) has gained enormous popularity in the neuroimaging community over the past few years. At the group level, most MVPA studies adopt an "information based" approach in which the sign of the effect of individual subjects is discarded and a non-directional summary statistic is carried over to the second level. This is in contrast to a directional "activation based" approach typical in univariate group level analysis, in which both signal magnitude and sign are taken into account. The transition from examining effects in one voxel at a time vs. several voxels (univariate vs. multivariate) has thus tacitly entailed a transition from directional to non-directional signal definition at the group level. While a directional group-level MVPA approach implies that individuals have similar multivariate spatial patterns of activity, in a non-directional approach each individual may have a distinct spatial pattern. Using an experimental dataset, we show that directional and non-directional group-level MVPA approaches uncover distinct brain regions with only partial overlap. We propose a method to quantify the degree of spatial similarity in activation patterns over subjects. Applied to an auditory task, we find higher values in auditory regions compared to control regions.

To achieve a certain sensory outcome, multiple actions can be executed. For example, unlocking a door might require clockwise or counterclockwise key turns depending on regional norms. Using fMRI in healthy human subjects, we examined the neural networks that dissociate intended sensory outcome from underlying motor actions. Subjects controlled a figure on a computer screen by performing pen traces on an MR-compatible digital tablet. Our design allowed us to dissociate intended sensory outcome (moving the figure in a certain direction) from the underlying motor action (horizontal/vertical pen traces). Using multivoxel pattern analysis and a whole-brain searchlight strategy, we found that activity patterns in left (contralateral) motor and parietal cortex and also right (ipsilateral) motor cortex significantly discriminated direction of pen traces regardless of intended direction of figure movement. Conversely, activity patterns in right superior parietal lobule and premotor cortex, and also left frontopolar cortex, significantly discriminated intended direction of figure movement regardless of underlying direction of hand movement. Together, these results highlight the role of ipsilateral motor cortex in coding movement directions and point to a network of brain regions involved in high order representation of intended sensory outcome that is dissociated from specific motor plans.

Abstract Evoked neural activity in sensory regions and perception of sensory stimuli are modulated when the stimuli are the consequence of voluntary movement, as opposed to an external source. It has been suggested that such modulations are due to motor commands that are sent to relevant sensory regions during voluntary movement. However, given the anatomical-functional laterality bias of the motor system, it is plausible that the pattern of such behavioral and neural modulations will also exhibit a similar bias, depending on the effector triggering the stimulus (e.g., right/left hand). Here, we examined this issue in the visual domain using behavioral and neural measures (fMRI). Healthy participants judged the relative brightness of identical visual stimuli that were either self-triggered (using right/left hand button presses), or triggered by the computer. Stimuli were presented either in the right or left visual field. Despite identical physical properties of the visual consequences, we found stronger perceptual modulations when the triggering hand was ipsi- (rather than contra-) lateral to the stimulated visual field. Additionally, fMRI responses in visual cortices differentiated between stimuli triggered by right/left hand. Our findings support a model in which voluntary actions induce sensory modulations that follow the anatomical-functional bias of the motor system.

No abstract

Summary The estimated accuracy of a classifier is a random quantity with variability. A common practice in supervised machine learning, is thus to test if the estimated accuracy is significantly better than chance level. This method of signal detection is particularly popular in neuroimaging and genetics. We provide evidence that using a classifier’s accuracy as a test statistic can be an underpowered strategy for finding differences between populations, compared to a bona fide statistical test. It is also computationally more demanding than a statistical test. Via simulation, we compare test statistics that are based on classification accuracy, to others based on multivariate test statistics. We find that the probability of detecting differences between two distributions is lower for accuracy-based statistics. We examine several candidate causes for the low power of accuracy-tests. These causes include: the discrete nature of the accuracy-test statistic, the type of signal accuracy-tests are designed to detect, their inefficient use of the data, and their suboptimal regularization. When the purpose of the analysis is the evaluation of a particular classifier, not signal detection, we suggest several improvements to increase power. In particular, to replace V-fold cross-validation with the Leave-One-Out Bootstrap.