Noriega, M;
Maranon, EJ;
Romero, D;
Orini, M;
Almeida, R;
(2016)
Respiratory rate estimation from multilead directions, based on ECG delineation.
In: Patton, J and Barbieri, R and Ji, J and Jabbari, E and Dokos, S and Mukkamala, R and Guiraud, D and Jovanov, E and Dhaher, Y and Panescu, D and Vangils, M and Wheeler, B and Dhawan, AP, (eds.)
2016 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC).
(pp. pp. 3813-3816).
IEEE
Preview |
Text
EMBC16_1243_FI.pdf - Accepted Version Download (134kB) | Preview |
Abstract
Estimating the instantaneous respiratory rate (Rr) from the electrocardiogram (ECG) is of interest as respiration direct measurement in clinical situations is often cumbersome. In this study, the Rr was estimated from the same Final Directions of maximum projection (FD) used for multi lead ECG automatic delineation. Power spectral analysis over the directions based on QRS complex main peak and T wave onset, peak and end spatial loops was used for Rr estimation. On a subset of the Physionet MGH/MF dataset, the proposed method yielded more accurate Rr estimates (minimum mean absolute error (MAE), 2.82 bpm) than the frequency tracking algorithm (minimum MAE, 4.53 bpm) and Fourier-based frequency estimation (minimum MAE, 4.94 bpm) using each lead alone, outperforming also the weighted multi-signal oscillator-based algorithm estimates for two or three lead (minimum MAE, 3.04 bpm). It was also shown that the FD of the three orthogonalized leads from Principal Component algorithm, improve the performance of Rr estimation.
Archive Staff Only
 |
View Item |
