Microcontroller-Based Online Nerve Parameter Estimation for Diagnosis of Healthy Subject Using Real-Time Nerve Conduction Study Signal Acquired Using Voltage Controlled Neurostimulator

The embedding systems play an inevitable role in healthcare informatics and medical diagnosis. This article presents the design of a microcontroller (μC)-based online parameter detection and diagnosis system using real-time nerve signals. In developing the system, a low cost and low power battery-based voltage controlled neurostimulator is designed to stimulate the targeted sites of underlying nerve. In the second phase, a signal conditioning circuit is designed to accurately process and obtain the recorded signals for parameter extraction. In offline mode, an optimum neural network structure is determined from MATLAB simulation using prediagnosed nerve signal dataset through cross-validation technique which is then implemented in programmable interface controller (PIC) μC (PIC18F45K22) for online estimation of parameter and diagnosis. Afterwards, nerve signals are collected from various subjects in laboratory following the guidelines of standard nerve conduction study (NCS) which are processed through embedded programmed μC. Experimental results and subsequent comparison with the state-of-the-art designs indicate that the proposed system is promising and reliable for online detection of the NCS parameters and subsequent diagnosis.