Modeling of Pain on a FPGA-based Neural Network

Miguel Angelo de Abreu de Sousa and Thiago Felipe de Jesus Torres


Neural Networks, Pattern Recognition, Modeling of Pain, Function Approximation


Although unpleasant, pain is a necessary sensation. It is helpful to prevent further injuries by inducing protective reflex, withdrawal movements and to serve as an alarm of health problems to the body. Due to such importance, several studies have been made in order to understand pain mechanisms. This work presents an artificial neural network for modeling of pain implemented on a Field Programmable Gate Array (FPGA). The model is based on the Gate Control Theory of Pain and, for the purpose of allowing hardware implementation of some specific features of the model, a new method for the approximation to the activation function of the neural network was elaborated. Such proposal combines piecewise first- and second-order approximations and it is also presented in this paper.

Important Links:

Go Back