Assessment of odor perception related to stimulation modes in a mock MRI scanner

Numerous studies investigated the central-nervous processing of olfactory information in humans, often based on functional MRI studies. However, when describing the olfactory perception, the course of olfactory sensations over time has received little attention although this constitutes a fundamental portion of the experimental paradigm. Hence, the aim of this study was to evaluate overall perception and the perception of odors over the time course of repeated odorous stimulation under different conditions of stimulus presentation.

Twenty-eight subjects with a normal sense of smell were asked to lie in a mock MRI scanner during the experiment. Three conditions of odor presentations were Tubing, (odors presented intranasally with separate tubing for each nostril), mask (odors presented to a mask covering the subject's nose) and vacuum (odors presented into the bore of the sham scanner). A pleasant odor (peach) and an unpleasant (fish) odor were presented using a computer-controlled olfactometer. Participants were asked to evaluate intensity and pleasantness of the odors. They also rated the comfortability of each presentation mode. For the vacuum and mask conditions, subjects were tracked with the time course of the odor intensity.

For all sessions, intensity was influenced by conditions (F = 29.6, p < 0.001, ηp² = 0.8, mask > tube > vacuum) and odors (F = 11.8, p = 0.003, ηp² = 0.4, fish > peach). Pleasantness was influenced by odors (F = 26.2, p = 0.001, ηp² = 0.593, peach > fish), comfortability was significantly influenced by conditions (F = 9.3, p = 0.002, ηp² = 0.5, vacuum > mask > tube). For peach odor, pleasantness was positively correlated with intensity (r = 0.3, p = 0.03) and negatively with fish odor intensity (r = -0.4, p < 0.001). Overall, the onset latency of intensity ratings (mean = 3.5 s, SD = 2.6) was significantly longer than the offset latency (mean = 1.6 s, SD = 0.8) (t = 3.5, p = 0.001). For the amplitude, conditions had a significant effect (F = 4.5, p = 0.04, ηp² = 0.2, mask > vacuum).

Perceived odor intensity was affected by the odor presentation being strongest in the mask condition, whereas this was not different for odor pleasantness. The vacuum condition was rated as most comfortable. Within the present design, the latency of onset and offset does not seem to be influenced by conditions, odors and blocks. Models of fMRI analyses should be adjusted to account for this shift in odor intensity perception.