Semantical Visual Information Facilitates Odor Imagery: A Combined Neurophysiological and Psychometrical Approach

Odor imagery (OI), or the ability to mentally simulate the presence of a smell, is a difficult cognitive function and is therefore misunderstood in terms of its neural underpinnings. In particular, the diverging results obtained in neuroimaging studies could be explained in part by the characteristics of the visual cues used to trigger this task. In this study, we investigated this question by comparing the effects of plain color patches, pictures, and words during OI using neurophysiological and psychometrical measurements.

Thirty healthy right-handed participants performed an OI task during a functional magnetic resonance imaging exam. They were instructed to focus on the odors evoked by different types of visual cues. Two 595s functional runs were performed using a block design in which seven individual plain colors, pictures, and odor labels were used as visual triggers for OI. After each functional run, behavioral ratings of success/failure of OI as well as subjective pleasantness and intensity of odor images were collected.

The OI task induced activity not only in regions of the olfactory network but also in non-olfactory brain regions. Across visual conditions, we observed activity in the orbitofrontal cortex, insula, supplementary motor area, and dorsolateral prefrontal cortex. Strikingly, no significant activity was found in the piriform cortex despite overall success in the OI task according to behavioral ratings (83% to 93% depending on the condition).

Our findings suggest that OI can occur without significant involvement of the primary olfactory regions, contradicting most of the literature on this topic. Only little behavioral and neurophysiological effects of visual conditions were revealed in our sample of good odor imagers, although semantical cues seem to require fewer cognitive resources to generate odor images. This assumption remains to be confirmed using specific tools for the measurement of cognitive load.