An enhanced design and control of a robotic hand using EEG signals

Abstract

Patients affected by Amyotrophic Lateral Sclerosis, Spino Cerebellar Ataxia, multiple newlinesclerosis, or strokes have limited mobility in the hospital bed or at home. A self-care newlinesystem which can communicate with the external world would help them to reduce newlinethe nursing load and to improve patients mobility and quality of life. Brain-Computer newlineInterface(BCI) is one such method which finds application in providing the real-time newlinesolutions. In BCI, Electroencephalography (EEG) is the most commonly used to extract the brain signal, which is easy to use, cost effective, and has fine resolution. newlineThis research work proposes a novel method for controlling the EEG based robotic hand using LabVIEW as a software environment. This implementation is carried out in three stages: the design of EEG acquisition system, development of the robotic hand and application of LabVIEW software as the interface between the two with audio buzzer as a neurofeedback. Here EEG acquisition circuit is designed, which incorporates advantageous features like low cost, good performance and easy operability using batteries as the power source. The acquisition circuit is tested against different forms of EEG signals like delta, theta, alpha, and beta and for different age groups. The requirements and dimensions of a human hand for different subjects were studied in detail in designing a robotic hand, the manipulators and the wrist design use articulated and spherical configuration methods, while the movement of fingers is based on a single tension cable method. At first, the EEG signals are extracted to control the closure and opening of robotic hand using motor imagery (thought process) signals. The extracted signal is then passed through the signal processing algorithms like Fast Fourier Transform (FFT) and Online Multi-resolutionWaveletAnalysis for Peak Detection (OM-WPD) and frequency detection of beta waves. The experimental results provide evidence that three different subjects in varying age groups had a similar FFT wave pattern in controlling.