Short- and medium-term product acceptance:;the neural substrates of taste acceptance in the temporal dimension, investigated with fMRI

The sense of taste is possibly the most elusive, and difficult to study.
Previous functional Magnetic Resonance studies mostly investigated the brain
activity related to tastes both 'extreme' and 'simple' , i.e. tastes obviously
either pleasant or unpleasant (sucrose, quinine), and not products one
encounters in real life: ingesting sucrose is different from tasting an apple
pie, or drinking apple juice. Whilst this is the sensible approach for
exploring a novel field, more ecologically valid tastes were left untested.
Moreover, the temporal evolution of the appreciation for a certain taste has
received scarce, if any, attention from the scientific community.
Understanding the neural substrates of the mechanisms behind the genesis of a
long-lasting appreciation of a certain taste would be crucial in the
formulation of foods, or food supplements, for strata of the population such as
children, elderly, cancer patients, all of them at risk of malnutrition
(especially under-nutrition). To address the clear difficulty of this study, we
plan to adopt both mainstream and advanced numerical techniques - such as
machine learning - to investigate and model the temporal evolution of the brain
response to tastes. The same techniques could then be applied to further
studies on the categories listed previously, namely elderly and cancer
patients.

The main objective of the study is to gain a deeper understanding of the
functioning of the human brain regarding the temporal evolution of the brain
activity associated with taste, specifically about taste acceptance.
In order to do so, we will develop a model capable to predict acceptance of a
taste on the short- and medium-term, based on the brain activity measured at
various time points, employing machine learning techniques. This model will
allow us to gain insights on the long-term time frame
The realization of that model, and the associated technique, will actually
constitute a second objective of the study. Such a model/technique could be
crucial in several other research fields, such as temporal evolution and
outcome of emotions, pain, and other stimuli.

The participants will be subjected to the following paradigm:

1) a screening visit, during which the experiment will be explained in detail,
and questions are answered. The participants will experience a training session
in a dummy MR scanner (an empty casing); this session will serve also to the
purpose to reduce the risk of claustrophobia in the real MR environment;

2) three scanning session, during which the paticipants will experience a
number of different tastes, administered in liquid form in small quantity per
administration (approx. 2 ml); upon receiving a taste, participants will rate
it for like/dislike by means of buttons.

3) the scanning sessions will happen in the morning, before breakfast, on an
empty stomach;

4) the second scanning session will take place one week after the first

5) participants will come back every day, to try and rate each product on a
discrete scale from -3 (max. dislike) to +3 (max. like) between the first and
the second session;

6) the tastes will be as follows: either four milk-based products and three
water-based, or eight Oral Nutritional Supplements.

The whole dataset will be then composed by behavioural data (the ratings
expressed during the scanning sessions, plus the ratings expressed between the
scanning sessions 1 and 2) as well as by brain activity data (recorded during
the two scanning sessions mentioned above).
Additionaly, an high-resolution anatomical MR scan will be used to match the
brain activity with brain structures at an individual level.

The analyisis of the dataset will include:

- standard preprocessing of the MR data: realignment, coregistration between
functional and anatomical data, spatial smoothing, normalization;

- single subject analysis of the brain activity, looking at the differences
between conditions (i.e. between the experience of different tastes), as well
as the differences between the experience of a taste and the baseline activity
of the brain (looking at a fixation cross); the results of the behavioural data
will be incorporated as covariates.

- application of machine learning techniques to the combined brain activity and
behavioural response data.

- group analysis of the output of the results collected at single subject
level, using both parametric and non parametric techniques

- The exposure to a Magnetic Resonance environment is not considered harmful
per se.
- The insurgence of a claustrophobic feeling is not uncommon; in order to
reduce this risk, we (a) ask the participant about being prone to
claustrophobia, and (b) put the participant in a dummy scanner (a far lesser
stressful environment) before to expose him to the real MR environment.
- During the scanning sessions the participants receive small amount of liquids
in their mouths (approx. 2 ml); this could provoke, in cases of mis-swallowing,
coughing. The small amount involved makes the insurgence of a life-threatening
condition unlikely to the extreme.

We believe that for the present study the ratio between the possible increase
of knowledge on the one hand and the burden/risk on the other hand is
favourable.
It is an acknowledged fact that Magnetic Resonance is an eminently risk-free
investigational tool.
The aim of the study is to investigate the evolution in time of the brain
activity that leads to the acceptance of a taste, which clearly rules out the
use of unpleasant tastes: it stands to reason, after all, that an awful taste
will be an unlikely candidate for long-term acceptance.
Conversely, this study promises to shed light on the inner mechanisms
subserving the evolution in time of how the human brain reacts to tastes; this
could, ultimately, lead to the development of new feeding strategies for strata
of the population where illness, or age are relevant factors as well as for
normal food consumers.
About the benefits for individual participants: there are no direct benefits.
However, they will benefit from an early detection if any structural anomaly is
detected at the level of their brain.