E Wellnhofer1, V Combé, H Oswald, E Fleck. 1. Charité, Campus Virchow Klinikum & Deutsches Herzzentrum Berlin, Humboldt University, Berlin, Germany. EWellnhofer@t-online.de
Abstract
UNLABELLED: Fluid-filled systems are generally used for invasive pressure measurements in cardiology, anesthesiology and intensive care medicine. Wave reflection and attenuation cause considerable signal distortion. METHODS: The transducer signal is amplified (no filtering) and sampled (rate 1 kHz) using an autocorrelation based algorithm to detect instantaneous cycle length. A digital Fourier transformation (DFT) for each heart cycle is performed. Amplitude and phase distortion are corrected using data matrices determined in in vitro experiments or calibration measurements for each fluid-filled system to be used. As a measure for accuracy the maximum of the difference of reference and corrected pressures (DIFF) was selected. 960 analyses were performed to assess the impact of correction, used system, mean pressure, time and A/D sampling rate on the agreement with reference pressure. Clinical examples are presented. RESULTS: Mean pressure was correlated with DIFF (r = 0.83). The correction algorithm achieves a significant (p < 0.001) reduction of DIFF from 20-30 mm Hg to 0-5 mm Hg in the high pressure range and from 1-3 mm Hg to 0-1.5 mm Hg in the low pressure system in in vitro experiments and in clinical pressure recordings. Sampling frequency < 1 kHz reduces accuracy. CONCLUSIONS: High fidelity correction of pressure signals from fluid-filled systems by harmonic analysis is feasible.
UNLABELLED: Fluid-filled systems are generally used for invasive pressure measurements in cardiology, anesthesiology and intensive care medicine. Wave reflection and attenuation cause considerable signal distortion. METHODS: The transducer signal is amplified (no filtering) and sampled (rate 1 kHz) using an autocorrelation based algorithm to detect instantaneous cycle length. A digital Fourier transformation (DFT) for each heart cycle is performed. Amplitude and phase distortion are corrected using data matrices determined in in vitro experiments or calibration measurements for each fluid-filled system to be used. As a measure for accuracy the maximum of the difference of reference and corrected pressures (DIFF) was selected. 960 analyses were performed to assess the impact of correction, used system, mean pressure, time and A/D sampling rate on the agreement with reference pressure. Clinical examples are presented. RESULTS: Mean pressure was correlated with DIFF (r = 0.83). The correction algorithm achieves a significant (p < 0.001) reduction of DIFF from 20-30 mm Hg to 0-5 mm Hg in the high pressure range and from 1-3 mm Hg to 0-1.5 mm Hg in the low pressure system in in vitro experiments and in clinical pressure recordings. Sampling frequency < 1 kHz reduces accuracy. CONCLUSIONS: High fidelity correction of pressure signals from fluid-filled systems by harmonic analysis is feasible.