A Real-Time ENG Pattern Recognition System based Sensory Event detection for Closed-Loop Control

Abstract

The goal of the present study was to detect real-time continuous sensory event from tactile afferent signals. To implement real-time continuous sensory event detection, we propose a novel electroneurogram (ENG) based pattern recognition. The tactile afferent signals were recorded from 16-channel micro-electrode on the dorsal root ganglion and feature vectors were extracted by a novel combination of time-frequency domain (TFD3). Principal component analysis (PCA) was used to reduce dimensionality of feature vectors. Finally, multilayer perceptron (MLP) classifier was used to detect sensory events. This result shows that the proposed method can be used for real-time continuous sensory event detection for closed-loop control.