NONLINEAR ANALYSES OF ELECTROGASTROGRAMS DURING OLFACTORY STIMULATION WITH VANILLA ODORANT, 152-158.

Eiji Takai, Kohki Nakane, and Hiroki Takada

Keywords

Electrogastrogram, Wayland algorithm, false nearest neighbour, olfactory stimulation, vanilla

Abstract

To examine the effect of odour on gastric motility, we took electrogastrogram (EGG) measurements during olfactory stimulation with different concentrations of vanilla odorants. Nonlinear analyses were used to examine EGG measurements. The translation error estimated by the Wayland algorithm increased proportionally to the concentration of vanilla odorants, suggesting that the translation error can be an indicator of odour intensity. The false nearest neighbour (FNN) method demonstrated that the minimum embedding dimension exhibited 5-6 dimensions, regardless of the odorant concentration; however, the odourless samples exhibited eight dimensions. Also, we estimated the minimum embedding dimension using a recently proposed method that does not require setting a threshold and uses the change rate in the nearest neighbour distance as an indicator; we found that the minimum embedding dimension exhibited two dimensions for each experimental group. These results suggest that olfactory stimulation affects the determinism of EGGs but not the minimum embedding dimension.

Important Links:

Go Back