Mental imagery of speech implicates two mechanisms of perceptual reactivation

Abstract

Sensory cortices can be activated without any external stimuli. Yet, it is still unclear how this perceptual reactivation occurs and which neural structures mediate this reconstruction process. In this study, we employed fMRI with mental imagery paradigms to investigate the neural networks involved in perceptual reactivation. Subjects performed two speech imagery tasks: articulation imagery (AI) and hearing imagery (HI). We found that AI induced greater activity in frontal-parietal sensorimotor systems, including sensorimotor cortex, subcentral (BA 43), middle frontal cortex (BA 46) and parietal operculum (PO), whereas HI showed stronger activation in regions that have been implicated in memory retrieval: middle frontal (BA 8), inferior parietal cortex and intraparietal sulcus. Moreover, posterior superior temporal sulcus (pSTS) and anterior superior temporal gyrus (aSTG) was activated more in AI compared with HI, suggesting that covert motor processes induced stronger perceptual reactivation in the auditory cortices. These results suggest that motor-to-perceptual transformation and memory retrieval act as two complementary mechanisms to internally reconstruct corresponding perceptual outcomes. These two mechanisms can serve as a neurocomputational foundation for predicting perceptual changes, either via a previously learned relationship between actions and their perceptual consequences or via stored perceptual experiences of stimulus and episodic or contextual regularity.

Keywords: Internal forward model/efference copy/corollary discharge; Memory retrieval; Mental simulation; Prediction; Sensorimotor integration.

Fig. 1

Dual stream prediction model (DSPM). Top: approximate cortical regions in the hypothesized dual streams. Bottom: schematic diagram of the DSPM (color scheme corresponds to the anatomical locations above). The abstract auditory representations (orange) can be induced via perception and perceptual reactivation and are formed around STG and STS. The perceptual reactivation process can be carried out in either the memory-retrieval or simulation-estimation prediction pathway. The memory-retrieval stream (blue) includes pMTG, MTL and distributed frontal-parietal networks for retrieval from long-term lexical items, episodic and semantic memory, respectively. The simulation-estimation stream (red) includes the frontal motor system and parietal somatosensory system. The articulatory trajectory is planned in frontal motor regions, including IFG, PMC, INS and SMA. If covert production is the goal, the planned articulation signal bypasses M1 and is simulated internally. The somatosensory consequence of the simulated articulation is estimated over parietal somatosensory regions, including SI, SII, PO and SMG. The auditory consequences – in the form of an abstract auditory representation – is derived from the subsequent estimation. A highly specified auditory representation (thick arrow) is generated in the bottom-up perceptual process that goes through spectrotemporal analysis of external stimuli in STG (brown). The stream containing the motor simulation and perceptual estimation processes can enrich the specificity of predicted auditory representations (solid arrows), compared to enrichment from the memory-retrieval stream (dotted arrows). Abbreviations: STG, superior temporal gyrus; STS, superior temporal sulcus; pMTG, posterior middle temporal gyrus; MTL, middle temporal lobe; IFG, inferior frontal gyrus; PMC, premotor cortex; INS, insula; SMA, supplementary motor area; M1, primary motor cortex; PO, parietal operculum; SI, primary somatosensory cortex; SII, secondary somatosensory cortex; and SMG, supramarginal gyrus.

Fig. 2

Functional MRI data for the AI and HI tasks. Statistical parametric t-maps indicate strength of the BOLD signal (p < .001 uncorrected) with cluster size greater than 25 voxels. (a) Shared cortical regions that mediate both AI and HI. (b) Regions show greater activity in HI compared to that in AI, including MFG, IPC and IPS, consistent with frontal-parietal distributed memory retrieval networks. (c) Regions show greater activity in AI compared to that in HI, including SMC, PO, which demonstrate the greater recruitment of simulation-estimation stream in AI. Greater activity is also observed in aSTG and pSTS, suggesting that more specific and robust auditory representation is obtained during AI. Abbreviations: SMA, supplementary motor area; IFG, inferior frontal gyrus; INS, insula; FO, frontal operculum; PO, parietal operculum; PMC, premotor cortex; MFG, middle frontal gyrus; IPC, inferior parietal cortex; IPS, intraparietal sulcus; SMC, sensorimotor cortex; aSTG, anterior superior temporal gyrus; pSTS, posterior superior temporal sulcus.

Fig. 3

Functional MRI data for common activation in AI and A. Statistical parametric t-maps indicate strength of the BOLD signal (p < .001 uncorrected) with cluster size greater than 25 voxels. Abbreviations: SMA, supplementary motor area; IFG, inferior frontal gyrus; INS, insula; ACC, anterior cingulate cortex; GP, globus pallidus; PO, parietal operculum.