“…In the first (training) phase, a new computational model of type-2 fuzzy regression (reasoning followed by defuzzification) is developed to fit the brain (hemodynamic) response of the subject to an aromatic stimulus as the input, and the oral response of the subject about the qualitative degree of concentration perceived by him/her in the form: Very Low, Low, Medium, High and Very High as the output. For convenience of realization, each qualitative grade is defined as sub-intervals of [0, 100], such as [1,20) for Very Low, [20,40) for Low, [40,60) for Medium, [60,80) for High, and [80, 100] for Very High. Intervals of width 20, instead of absolute value in [0, 100] is utilized to express subject's perceptual response to avoid extensive subject's training for each degree of concentration in [0, 100].…”