Motion artifact contaminated multichannel EEG dataset

Wearable EEG suffers from motion artifact contamination due to the subject's movement in an ambulatory environment. Signal processing techniques pose promising solutions for the detection and removal of motion artifacts from ambulatory EEG, but relevant open-access datasets are not available, which is detrimental to the development of wearable EEG applications. This article showcases open-access electroencephalography (EEG) recordings, while a subject is performing different upper-body, lower-body, and full-body movements. One healthy male subject volunteered to record his EEG data using a 14-channel EMOTIV EPOCH EEG headset device. This device's electrode placement is in accordance with the international 10-20 system, and the data was stored using the EMOTIV Pro application. We used the MATLAB software to visualize the captured brain waveforms. The venue of the data collection was the Biomedical Instrumentation and Signal Processing Laboratory (BISPL) at the Independent University, Bangladesh (IUB). The EMOTIV Pro application extracted the recorded EEG data in the CSV file format, while the MATLAB software converted it to a .mat extension file afterward. The first 14 columns of this file represent the 14-channel EEG data, and the subsequent nine columns are for the motion sensor data. The list of recorded movements includes blinking of eyes, eyebrow movement, and also horizontal and vertical eye movements. Afterward, the head shook and nodded. Later, the leg trembled, followed by listening to music, talking, walking, and standing and sitting down. Before the recording ended, the subject relaxed on a chair with both eyes open and closed. This dataset is one of its kind, allowing us to explore further research for wearable EEG while denoising motion artifacts arising from subject movement.