EEG are surely prone to be contaminated with 'noise', i.e., signals not related to brain activity. This include EMG from muscles attached to the cranium (including facial muscles, e.g., as evident during eye blinking) as well as eye movements (the eyes are significant dipoles and eye movements therefore strongly affects EEG recordings). In addition, it's difficult to ensure stable electrode impedance and ambient electromagnetic noise.
\n\nThe data set contains noise of these kinds although elaborate precautions were take to minimize them including performing the recordings in a shielded room (Faraday cage) and designing a task that required minimal head and torso movements. In other words, the signal quality of the data set - despite the 'noise' - is superior to anything that can be expected under 'real-life' scenarios (e.g., when riding a wheel chair).
\n\nGuocong Song raised concerns regarding the sensor system system (Polhemus FASTRAK) used for recording the 3D positions of the object, the hand and the engaged digits. Polhemus uses a stationary signal source that generates short-lasting electromagnetic pulses (@~10 kHz) alternating between 3 planes (X, Y, Z) at a rate of 120 Hz. In theory these pulses could be demodulated by EEG electrodes (giving rise to 'pulses' with 120 and 360 Hz frequency). Probably needless to say, no overt evidence of interference from the Polhemus were observed during the experiments, but it is possible that it was present in the recordings. Importantly, the EM source was stationary throughout the experiments and the fact that the sensors themselves were moving in space is of no consequence.
", "messages": [ { "id": 86654, "postDate": "2015-07-23T15:29:10.007Z", "content": "EEG are surely prone to be contaminated with 'noise', i.e., signals not related to brain activity. This include EMG from muscles attached to the cranium (including facial muscles, e.g., as evident during eye blinking) as well as eye movements (the eyes are significant dipoles and eye movements therefore strongly affects EEG recordings). In addition, it's difficult to ensure stable electrode impedance and ambient electromagnetic noise.
\n\nThe data set contains noise of these kinds although elaborate precautions were take to minimize them including performing the recordings in a shielded room (Faraday cage) and designing a task that required minimal head and torso movements. In other words, the signal quality of the data set - despite the 'noise' - is superior to anything that can be expected under 'real-life' scenarios (e.g., when riding a wheel chair).
\n\nGuocong Song raised concerns regarding the sensor system system (Polhemus FASTRAK) used for recording the 3D positions of the object, the hand and the engaged digits. Polhemus uses a stationary signal source that generates short-lasting electromagnetic pulses (@~10 kHz) alternating between 3 planes (X, Y, Z) at a rate of 120 Hz. In theory these pulses could be demodulated by EEG electrodes (giving rise to 'pulses' with 120 and 360 Hz frequency). Probably needless to say, no overt evidence of interference from the Polhemus were observed during the experiments, but it is possible that it was present in the recordings. Importantly, the EM source was stationary throughout the experiments and the fact that the sensors themselves were moving in space is of no consequence.
", "rawMarkdown": "\r\nEEG are surely prone to be contaminated with 'noise', i.e., signals not related to brain activity. This include EMG from muscles attached to the cranium (including facial muscles, e.g., as evident during eye blinking) as well as eye movements (the eyes are significant dipoles and eye movements therefore strongly affects EEG recordings). In addition, it's difficult to ensure stable electrode impedance and ambient electromagnetic noise. \r\n\r\nThe data set contains noise of these kinds although elaborate precautions were take to minimize them including performing the recordings in a shielded room (Faraday cage) and designing a task that required minimal head and torso movements. In other words, the signal quality of the data set - despite the 'noise' - is superior to anything that can be expected under 'real-life' scenarios (e.g., when riding a wheel chair). \r\n\r\nGuocong Song raised concerns regarding the sensor system system (Polhemus FASTRAK) used for recording the 3D positions of the object, the hand and the engaged digits. Polhemus uses a stationary signal source that generates short-lasting electromagnetic pulses (@~10 kHz) alternating between 3 planes (X, Y, Z) at a rate of 120 Hz. In theory these pulses could be demodulated by EEG electrodes (giving rise to 'pulses' with 120 and 360 Hz frequency). Probably needless to say, no overt evidence of interference from the Polhemus were observed during the experiments, but it is possible that it was present in the recordings. Importantly, the EM source was stationary throughout the experiments and the fact that the sensors themselves were moving in space is of no consequence. ", "votes": 3 } ], "comments": [], "raw_markdown_by_id": { "86654": "\r\nEEG are surely prone to be contaminated with 'noise', i.e., signals not related to brain activity. This include EMG from muscles attached to the cranium (including facial muscles, e.g., as evident during eye blinking) as well as eye movements (the eyes are significant dipoles and eye movements therefore strongly affects EEG recordings). In addition, it's difficult to ensure stable electrode impedance and ambient electromagnetic noise. \r\n\r\nThe data set contains noise of these kinds although elaborate precautions were take to minimize them including performing the recordings in a shielded room (Faraday cage) and designing a task that required minimal head and torso movements. In other words, the signal quality of the data set - despite the 'noise' - is superior to anything that can be expected under 'real-life' scenarios (e.g., when riding a wheel chair). \r\n\r\nGuocong Song raised concerns regarding the sensor system system (Polhemus FASTRAK) used for recording the 3D positions of the object, the hand and the engaged digits. Polhemus uses a stationary signal source that generates short-lasting electromagnetic pulses (@~10 kHz) alternating between 3 planes (X, Y, Z) at a rate of 120 Hz. In theory these pulses could be demodulated by EEG electrodes (giving rise to 'pulses' with 120 and 360 Hz frequency). Probably needless to say, no overt evidence of interference from the Polhemus were observed during the experiments, but it is possible that it was present in the recordings. Importantly, the EM source was stationary throughout the experiments and the fact that the sensors themselves were moving in space is of no consequence. " } }