Shared Neural Mechanisms of Visual Perception and Imagery.

TLDR: There is a large overlap in neural processing during perception and imagery: neural representations of imagined and perceived stimuli are similar in the visual, parietal, and frontal cortex, and perceptions and imagery seem to rely on similar top-down connectivity.

About: This article is published in Trends in Cognitive Sciences.The article was published on 2019-05-01 and is currently open access. It has received 143 citations till now. The article focuses on the topics: Visual perception & Mental image.

Figures

Figure 3. Temporal dynamics during perception and imagery. Temporal generalization matrices: a classifier is trained on a specific time point and then tested on all other time points, providing a measure of how well neural representations at certain time points generalize. The y-axis represents the training time and the x-axis represents the testing time. The colors show decoding accuracy. (A) Adapted from [52]. Temporal generalization of MEG signals during object perception. The eclipse shows diagonal generalization, indicating rapidly changing representations. The rectangle indicates off-diagonal generalization indicating a stable representation around that training time. (B)-(C) Adapted from [54] (B) Temporal generalization of MEG signals during imagery of faces and houses. The square pattern indicates broad off-diagonal generalization. (C) Temporal generalization of MEG signals between perception and imagery. A classifier was trained at different time points during imagery and tested at different time points during perception. Perceptual processing around 130 ms and after 300 ms generalized to imagery.

Figure 1. Overlap in neural representations between perception and imagery estimated using fMRI. (A) Adapted from [14]. The overlap in neural representations of imagined and perceived stimuli was assessed using cross-decoding in four different regions of interest along the ventral visual stream. The results indicate that there is more overlap between imagery and perception in high-level, object-selective areas compared to low-level, retinotopic areas. (B) Adapted from [14]. The overlap in retinotopic areas was correlated with participants’ imagery vividness scores. The positive correlation shows that people reporting more vivid imagery have higher overlap with perception in these areas. (C) Adapted from [25]. The neural overlap between imagined and perceived stimuli was assessed across the whole brain. Red/yellow indicates significant overlap and blue/green indicates a modulation of the overlap by imagery vividness. There was significant overlap in visual, parietal and frontal areas but only the overlap in visual areas is modulated by imagery vividness.

Figure 4. Effective connectivity during perception and imagery. Adapted from Dijkstra et al. (2017b). The results were obtained using Dynamic Causal Modelling (DCM) of fMRI data during perception and imagery on four ROIs: early visual cortex (EV), high-level visual cortex/fusiform gyrus (FG), intraparietal sulcus (IPS) and inferior frontal gyrus (IFG). The shown arrows reflect connections that were significantly modulated on the group level by the different factors. The width of the arrows indicate the strength of the influence and light arrows reflect inhibitory connections while dark arrows reflect excitatory connections. Both imagery and perception rely on top-down recruitment of visual areas. The vividness of visual imagery correlates with the strength of top-down recruitment of early visual areas. Furthermore, whereas perception is accompanied by bottom-up connectivity, this is absent during imagery.

Figure 2, Key Figure. Shared neural mechanisms of visual perception and imagery. Neural representations of imagined and perceived stimuli show overlap in ventral visual areas as well as parietal and frontal areas. During both perception and imagery, visual areas represent the visual features of the sensory experience. Parietal areas encode the saliency, which is used for top-down spatial and featured-based attention during perception as well as during imagery. Representations in frontal areas are more task- than content-dependent, indicating that frontal areas represent task-relevant structure during both imagery and perception. There seems to be large overlap in top-down mechanisms between imagery and perception whereas the bottom-up processing characteristic of perception is absent during imagery.

Frequently asked questions

Q1. What future works have the authors mentioned in the paper "Shared neural mechanisms of visual perception and imagery" ?

Future research should explore the function of each of these representations within this simulation process. To explore this issue further, future research could investigate how neural representations change when people are aware that what they see is not there, such as during voluntary imagery, versus when they are not aware, such as during hallucinations. Future research should explore to what extent these conclusions extend to other modalities ( see Box 3 ). 

Q2. What have the authors contributed in "Shared neural mechanisms of visual perception and imagery" ?

Here, the authors review recent neuroimaging studies comparing these two forms of visual experience. Their results suggest that there is large overlap in neural processing during perception and imagery: neural representations of imagined and perceived stimuli are similar in visual, parietal and frontal cortex. Furthermore, perception and imagery seem to rely on similar top-down connectivity. 

Q3. What is the main hypothesis for the coupling between IPS and visual cortex?

One hypothesis is that this couplingreflects the transformation from more abstract stimulus information into more sensory stimulus representations. 

Q4. What is the significance of top-down coupling between IFG and visual cortex?

Top-down coupling between IFG and visual cortex has been proposed to be important for selective attention in the presence of a visual stimulus [75,76], as well as for the maintenance of visual information in the absence of a stimulus [37,77]. 

Q5. What is the effect of DCM on visual cortex?

Using DCM on fMRI data [72,73], a recent study showed that during both imagery and perception, there is a strong increase in top-down coupling between inferior frontal gyrus (IFG) and visual cortex compared to baseline [74] (Fig. 4). 

Q6. What is the effect of the top-down coupling between IPS and visual cortex?

During imagery, this coupling is stronger, which might lead to increased activation of these sensory representations and therefore give rise to internally generated visual experience in the absence of bottom-up input. 

Q7. How long does imagery overlap with perceptual processing?

The results indicate that imagery specifically overlaps with perceptual processing around 150 milliseconds and after 300 milliseconds. 

Q8. What is the evidence for recruitment of the corresponding sensory cortex during auditory imagery?

There is evidence for recruitment of the corresponding sensory cortex during auditory imagery [104,105], odor imagery [106,107], tactile imagery [108] and motor imagery [109,110].