Multichannel filter for heartbeat detection in noisy ECG recordings.

In ECG signals recorded with smart clothes disturbances as intermittent loss of signal from electrodes, movement artefacts, and electromyographic interference are common. In this study a multichannel method for spatio-temporal filtering is evaluated using ECG signals from a database and with three recordings made with a T-shirt with integrated textile electrodes. The sensitivity and precision of the signals from the database were 99.6% and 98.5%, respectively, if 12 channels were used and the signal-to-noise ratio was -10 dB. The filter gave a sensitivity of 99.6% and a precision of 99.5% in the recordings from the textile electrodes. In conclusion, the results obtained indicated that multichannel spatio-temporal filtration could be a suitable method for heartbeat detection in ECG measurements with textile electrodes. [ABSTRACT FROM AUTHOR]