Stimulus level estimation using functional near-infra-red data and neural network

In this paper we applied a well-tested neural network, so-called Supervised Fuzzy Adaptive Resonance (SF), to investigate the potential of functional Near-Infra-Red (fNIR) spectroscopy for automated assessment of physical stimulus intensity. To induce mild, moderate and severe physical stimuli, we asked the participants to keep their left hand in the ice water for gradually increased durations. Initial tests with fNIR data from 6 healthy participants (36 trials) indicated that SF is a reliable automated method to estimate the intensity of the induced stimuli with a high accuracy.