Examining the Software Implementation of Pain Sensation on the Fpga Board Using the Neuromorphic Concept

Abstract

Background: The sense of touch plays an important role in our interactions with the surrounding environment. Mechanical arms, robots and neural prostheses will function better with a sense of touch. Microneurography studies in humans have shown that primary afferent neurons such as receptors Fingertip mechanics play an important role in encoding and discriminating different types of stimuli using spike train patterns.Objective: To investigate the software implementation of pain sensation on FPGA board using the concept.Research method: a laboratory method for simulating the responses of the slow matching receptor. It has been applied to force stimulation by considering the spiking behavior. In fact, to detect the force, Izhikevich sensor data and spiking characteristics were used. The sensor's analog signal was applied as an input current to the neuron model in order to obtain spike trains. The features of the spike trains were extracted by rate coding and spike timing coding. ) and kNN (k-nearest neighbor) were given to classify all types of forces. Results: The highest accuracy of the rate coding feature class with 100% accuracy, (ISI CV) Inter-spike intravel coefficient of variation with 81.18% accuracy and (VPD) Victor-purpura distance with 82% accuracy was obtained. Also, rate coding and time coding were also calculated using spike trains resulting from mutual information contact force.Results: Sending information by rate coding method is more than spike timing coding in stimulating the Merkel receptor. Also, with increasing power, the firing rate of the Merkel receptor also increases.