Evaluating Dry Systems and Data Quality
Quantifying signal quality from an ECG or EEG system has always been challenging due to the lack of objective standards, the semi-random nature of the source signal and the inherent variability in human subjects. Dry systems add additional factors related to the sensor and mechanics and pure electronics specifications do not adequately cover the effects of real-world sensor-to-skin contact noise. Ultimately the question of whether or not a dry system is ‘good’ is answered by seeing if the dry system can deliver the same signal as the wet system.
In our experience, we have determined that it is necessary to evaluate dry systems on two levels. The first level involves usability – e.g., does the EEG headset fit on different heads with sensor contact across the scalp? Unlike wet systems, contact onto different head shapes is not a given since there are no gels to fill gaps between harness and scalp.
Simultaneous recording between wet (Ag/AgCl conductive gel) and dry electrodes (Flex, through hair) showing the high signal quality of Cognionics sensor systems. Bipolar montage to avoid the influence of noise from the mastoid reference.
The second level involves examining the signal from the dry sensor against the current gold standard – a wet sensor. In the above dataset, data was recorded at the exact same time from our dry Flex sensors and gel Ag/AgCl sensors. To simulate a wet sensor overlaid directly on a dry sensor, two wet sensors were placed adjacent to the dry sensor (while taking care not to contaminate the gel) and the wet signals were averaged. The example demonstrates the high raw signal quality of the Cognionics dry system as the wet and dry signals are virtually indistinguishable.
Evoked potentials (P300 and AEP) measured with both wet and dry systems showing very high correlation in the time averaged response.
For a more quantitative test, the above example measures the correlation between time averaged evoked responses from both wet and dry systems. The close agreement (r>0.9) shows that the dry system measures the same signal as wet electrodes. Furthermore, the test validates the timing accuracy of Cognionics wireless triggering system, compared to the wet system which utilized a standard wired connection.
For more information regarding dry sensors and the technology, please refer to our peer reviewed publications and papers.