EEG-based functional connectivity and physiological correlates of emotion in fragrance exposure
Podium 35
Presented by: Hugo Ferreira
Sensorial evaluation of a cosmetic product is a major aspect in product release to the public. The sensorial characteristics can be evaluated in naive or trained panels that simulate the behavior of future consumers. Nonetheless, self-reporting of the panels’ sensorial experience, as is conventionally done, rarely captures the emotional content with fidelity. This results in a limited translation to the consumer’s preferences.
The use of physiological responses, such as eye movement, facial mimicry, heart rate, skin conductance, and brain electrical activity, are ever more being used as more objective and quantitative correlates of salience and emotional content of subjective cosmetic product experience.
Herein, physiological correlates were assessed when testing a panel of fragrances, in order to define a normative emotional response according to the dimensions of valence, pleasure and dominance.
Four women, aged between 20-45, were enrolled in this study. Participants were blindly exposed to a panel of 7 different fragrances in a randomized order, including lavender, lemon, and vanilla scents, considered as “pleasant”, kerosene and “horse sweat” considered as “unpleasant”, and cedar and vinegar considered as “neutral”. A block design protocol was considered for the presentation of such fragrances and self-reporting feedback using the 3-dimensional Self-Assessment-Manikin (SAM). Instructions were presented visually in iMotions platform following the sequence: (1) 10-second exposure of the fragrances; (2) 20 seconds for assessing emotional response (3) 30 seconds for providing SAM feedback using a computer application (4) 30 seconds for scent wash-off, in which the participants were asked to smell their hands before the next fragrance. The first fragrance exposure was preceded by a 30-second baseline with no fragrance exposure. In the end of the exposures, participants were further asked to describe the emotions, feelings and thoughts elicited by each fragrance as they recalled them, as well as provide feedback on the general experience.
In parallel, whilst participants followed the above-mentioned protocol, their physiological data was being collected with a number of different devices interfacing with the iMotions platform. A Tobii eye tracker was used to monitor eye movement to assess salience to visual instructions and the behavior whilst self-reporting the SAM. A Logitech webcam was used to capture video of the participants’ faces and assess emotion content through Ekman’s facial mimicry model. A Schimer device was used to measure sympathetic activity through photoplethysmography-derived heart rate and galvanic skin response. Finally, a 64-channel Brain Vision actiCHamp Plus electroencephalography (EEG) system was used to measure electrical brain activity from which alpha (8-13 Hz) and theta (4-8 Hz) waves and functional connectivity metrics, such as phase lag index (PLI), were derived.
Data analysis was done using the iMotions platform and the MNE tools for EEG functional connectivity data. Finally, group statistical analysis was done using SPSS.
As expected, the exposure to fragrances elicited higher connectivity in brain regions belonging to the olfactory network such as the entorhinal cortex which receives direct olfactory bulbar input, and the insula as a secondary associative region related to salience and arousal. Additional limbic structures such as the parahippocampal gyrus and the cingulate were observed to have higher connectivity, which are involved in memory and emotional processes.
Pleasant fragrances were observed to elicit higher connectivity in the left hemisphere in contrast with unpleasant fragrances which elicited higher connectivity in the right hemisphere.
In particular, pleasant fragrances were associated with higher alpha connectivity in the left lateral-orbitofrontal cortex, a region involved in emotion response and which connectivity has been shown to be altered in depression.
Unpleasant fragrances were associated with higher alpha connectivity in the right pars triangularis, a region that is related to language.
Neutral fragrances on the other hand were associated with higher alpha connectivity in the left banks of superior temporal sulcus, which has been observed to be active during tasks of cognitive empathy and perspective-taking.
These results correlate with higher valences reported for pleasant fragrances, whilst unpleasant fragrances were described as eliciting higher arousal levels and to superimpose the participants’ sense of dominance.
As a resume, results suggest that EEG-based functional connectivity and physiological correlates may be able to provide a means to define objective gold-standards of emotional response for global marketing efforts in cosmetic products.
The use of physiological responses, such as eye movement, facial mimicry, heart rate, skin conductance, and brain electrical activity, are ever more being used as more objective and quantitative correlates of salience and emotional content of subjective cosmetic product experience.
Herein, physiological correlates were assessed when testing a panel of fragrances, in order to define a normative emotional response according to the dimensions of valence, pleasure and dominance.
Four women, aged between 20-45, were enrolled in this study. Participants were blindly exposed to a panel of 7 different fragrances in a randomized order, including lavender, lemon, and vanilla scents, considered as “pleasant”, kerosene and “horse sweat” considered as “unpleasant”, and cedar and vinegar considered as “neutral”. A block design protocol was considered for the presentation of such fragrances and self-reporting feedback using the 3-dimensional Self-Assessment-Manikin (SAM). Instructions were presented visually in iMotions platform following the sequence: (1) 10-second exposure of the fragrances; (2) 20 seconds for assessing emotional response (3) 30 seconds for providing SAM feedback using a computer application (4) 30 seconds for scent wash-off, in which the participants were asked to smell their hands before the next fragrance. The first fragrance exposure was preceded by a 30-second baseline with no fragrance exposure. In the end of the exposures, participants were further asked to describe the emotions, feelings and thoughts elicited by each fragrance as they recalled them, as well as provide feedback on the general experience.
In parallel, whilst participants followed the above-mentioned protocol, their physiological data was being collected with a number of different devices interfacing with the iMotions platform. A Tobii eye tracker was used to monitor eye movement to assess salience to visual instructions and the behavior whilst self-reporting the SAM. A Logitech webcam was used to capture video of the participants’ faces and assess emotion content through Ekman’s facial mimicry model. A Schimer device was used to measure sympathetic activity through photoplethysmography-derived heart rate and galvanic skin response. Finally, a 64-channel Brain Vision actiCHamp Plus electroencephalography (EEG) system was used to measure electrical brain activity from which alpha (8-13 Hz) and theta (4-8 Hz) waves and functional connectivity metrics, such as phase lag index (PLI), were derived.
Data analysis was done using the iMotions platform and the MNE tools for EEG functional connectivity data. Finally, group statistical analysis was done using SPSS.
As expected, the exposure to fragrances elicited higher connectivity in brain regions belonging to the olfactory network such as the entorhinal cortex which receives direct olfactory bulbar input, and the insula as a secondary associative region related to salience and arousal. Additional limbic structures such as the parahippocampal gyrus and the cingulate were observed to have higher connectivity, which are involved in memory and emotional processes.
Pleasant fragrances were observed to elicit higher connectivity in the left hemisphere in contrast with unpleasant fragrances which elicited higher connectivity in the right hemisphere.
In particular, pleasant fragrances were associated with higher alpha connectivity in the left lateral-orbitofrontal cortex, a region involved in emotion response and which connectivity has been shown to be altered in depression.
Unpleasant fragrances were associated with higher alpha connectivity in the right pars triangularis, a region that is related to language.
Neutral fragrances on the other hand were associated with higher alpha connectivity in the left banks of superior temporal sulcus, which has been observed to be active during tasks of cognitive empathy and perspective-taking.
These results correlate with higher valences reported for pleasant fragrances, whilst unpleasant fragrances were described as eliciting higher arousal levels and to superimpose the participants’ sense of dominance.
As a resume, results suggest that EEG-based functional connectivity and physiological correlates may be able to provide a means to define objective gold-standards of emotional response for global marketing efforts in cosmetic products.