Visual mental imagery engages the left fusiform gyrus, but not the early visual cortex: a meta-analysis of neuroimaging evidence

Abstract

The neural bases of visual mental imagery (VMI) are the object of intense debate. Their identification is central to define the brain substrates our conscious experience and can be clinically important to reveal consciousness in non-communicating patients or to alleviate posttraumatic stress disorder. The dominant model of VMI stipulates a functional and anatomical equivalence between visual mental imagery and visual perception. In patients with acquired brain damage, the model predicts a systematic co-occurrence of perceptual and imaginal deficits.However, patients with lesions restricted to the occipital cortex often experience vivid mental images, even in case of cortical blindness resulting from bilateral V1 damage. Instead, patients with extensive left temporal damage are more likely to have impaired VMI. On the other hand, some neuroimaging and neuromodulatory evidence does suggest an implication of striate cortex in VMI. To address this discrepancy, we conducted an activation-likelihood-estimation-based large-scale meta-analysis of 46 functional magnetic resonance imaging experiments to map the extent of cortical activation associated with three contrasts: (1) all studies combined; (2) VMI versus Control, (3) VMI versus Perception, and (4) Motor Mental Imagery versus Control.Results from the VMI versus Control contrast demonstrated an association between VMI and activation increase in the frontoparietal and cingulo-opercular networks, as well as in the left fusiform gyrus. Results from the VMI versus Perception contrast showed the association between VMI and activation increase of the anterior insular and supplementary motor area / anterior cingulate cortices, bilaterally. Conjunction analyses between the VMI versus Control and the VMI Mental Imagery versus Perception contrasts showed the activation of the left anterior insular cortex and in the bilateral supplementary motor area / anterior cingulate cortex.Results from the Motor Mental Imagery versus Control contrast showed that mental motor imagery increases the activation of the cerebellum and of the superior frontal gyri bilaterally, of the left supplementary motor area, of the anterior portion of the left calcarine sulcus, and of the left fusiform / parahippocampal gyri. Conjunction analyses between the VMI versus Control and the Motor Mental Imagery versus Control contrasts showed the activation of the bilateral superior frontal gyrus as well as of the left supplementary motor area / anterior cingulate cortex.Thus, the results stress the importance for VMI of brain networks associated with attentional control and working memory functions, together with rostral portions of the cortical ventral visual stream. Bayesian analysis confirmed the lack of evidence for an activation of the early visual areas in VMI, consistent with the results from brain-damaged patients. Our evidence suggests a revision of the VMI model. A Fusiform Imagery Node (FIN) in the area FG3 of the left fusiform gyrus might act as a hub retrieving visual information from long-term sem...

Untying the knot: imagination, perception and their neural substrates

“…But we might worry that they fail to pinpoint specific failings in this area. Spagna et al (2021) approach the V1 controversy from a different angle. They are concerned to highlight observations of brain-damaged patients in clinical contexts that are "sharply discordant" (p. 202) with the experimental results yielded from the study of healthy, neurotypical subjects, and which supposedly show the necessity of V1 for imagery.…”

“…Two kinds of vision/imagery double dissociations found in the clinical literature bear on this discussion. Reflection on them can be used to flesh out the above claims by Spagna et al (2021).…”

Modelling perception as a hierarchical competition differentiates imagined, veridical, and hallucinated percepts

“…Individuals still report being able to create mental images despite near-complete bilateral lesions to primary visual cortex, suggesting that the primary visual cortex is not necessary for visual mental imagery (Bridge et al, 2012;Chatterjee & Southwood, 1995;de Gelder et al, 2015;Zago et al, 2010). Further, a systematic review of the neural correlates of visual mental imagery by Spagna et al (2020) found that the left fusiform gyrus, a high-level region within the visual ventral stream, was the only region reliably involved during mental imagery across studies (although the authors do not exclude the involvement of more anterior regions of the temporal lobe given the known difficulty in imaging these regions). These findings are entirely in line with the predictions of our model: while primary sensory cortices can be involved in mental imagery, the only essential component of the experience is the internal generation of a non-veridical stimulus representation, which in our model occurs at the hierarchical apex, corresponding, for instance, to late in the visual ventral stream.…”

“…Future studies may therefore seek to expand the approach here to alternate models of sensory processing, although our model is still capable at emulating findings on how imagined and veridical information is distributed in real human brains (e.g. Lee et al, 2012;Spagna et al, 2020;VanRullen & Reddy, 2019).…”