This study performed observation of human brain activity in response to audio stimuli using Electrical Capacitance Volume Tomography (ECVT). The observation was done by interpreting and examining 32-channels ECVT Brain Scanner's image. Hearing activity in brain was recorded while a single tone stimulus with three different frequencies -500 Hz, 1000 Hz, and 4000 Hz, was played into the right ear of each subjects. ECVT's measurement data then being reconstructed into permittivity distribution-displayed as brain image, by using average subtraction method and ILBP algorithm. Since ECVT has limitation on its spatial resolution and its original reconstructed image of brain cannot specifically show the exact location of the activated brain yet, it needs further processing. Image is processed by determining centroid distance from entire produced frames to get sliced area which is estimated as auditory cortex - activated brain area when human hears. By this method, we can get a sliced area with permittivity distribution reach maximum value, 1. The sliced area then being segmented by its colour on a scale range at 0.8-1 - shown by red colour. Images processed by centroid method are correlated with coefficient of correlation is higher than 0.5 in all stimulus conditions. It shows that ECVT can be considered as a promising technique for classification of activated brain area for neuroscience applications.