Artificial Intelligence enabled neuroproteins design for Brain mapping dynamics based on motor imagery classification using HCI (Human computer interface) and (EEG) electroencephalogram.

Autor: Singh, Anuj, Sharma, Sachin, Purohit, Kamlesh Chandra
Předmět:
Zdroj: Library of Progress-Library Science, Information Technology & Computer; Jul-Dec2024, Vol. 44 Issue 3, p12354-12370, 17p
Abstrakt: The brain activity can be represented by EEG, which does it by measuring voltage fluctuations in the brain. This paper aims to formulate a system which is capable of classifying and predicting the class of motor movement corresponding to the left or right direction. Such systems can have applications in various fields of human work. There are two datasets used in this study, relating to motor movements - left/right hand movement and eye open/close. The left/right hand movement dataset used in this paper captured brain wave readings recorded with an equipment of 128 Hertz sampling frequency with 16-bit analog converter with 14 electrode channels. The collected data is pre-processed by applying Fast Fourier Transform (FFT) to obtain a frequency domain matrix. The result computed by applying FFT provides a condensed representation of the data. The data was recorded across a 25-minute period by showing participants visuals of left and right movements, and brain waves were recorded corresponding to each movement. The second dataset captured brain impulses corresponding to two states of the eyes - eyes opened and eyes closed, over a period of two minutes. In this work how the split signals were employed, and how they vary from person to person. This Designed system effectively separate and utilize signals, and the adoption of AI techniques in these devices appears to be the most viable option for society 5.0. The integration of neurological data collected from the brain is used to train the prosthesis to conduct activities without encountering any impediments in this signal. The overall efficiency of this system is 80 to 87%. The proposed AI enabled Neuroprosthetics are cost effective and user friendly. There is no need to have surgical implant to acquire the EMG signals from the individuals to operate Neuroprosthetics. [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index