TY - GEN
T1 - Stationary Wavelet Transform for the Extraction of the Impedance Circulation Component during Out-of-hospital Cardiac Arrest
AU - Isasi, Iraia
AU - Alonso, Erik
AU - Irusta, Unai
AU - Aramendi, Elisabete
AU - Daya, Mohamud R.
N1 - Funding Information:
This work was supported partially by the Spanish Ministry of Science, Innovation and Universities through Grant RTI2018-101475-BI00, jointly with the Fondo Europeo de Desarrollo Regional (FEDER), and in part by the Basque
Publisher Copyright:
© 2020 Creative Commons; the authors hold their copyright.
PY - 2020/9/13
Y1 - 2020/9/13
N2 - An automated pulse detector during out-of-hospital cardiac arrest (OHCA) is needed. The thoracic impedance (TI) recorded through defibrillation pads presents an impedance circulation component (ICC), hidden among other components, in the form of small fluctuations correlated with each effective heartbeat. This study pre-sentes a method based on the stationary wavelet transform (SWT) to derive the ICC. A dataset with 456 5-s segments, 175 pulseless electrical activity (PEA) and 281 pulse-generating rhythm (PR), with concurrent ECG and TI signals from 49 OHCA patients was used. The SWT was used to decompose the TI into 7 levels. The ICC was derived from soft denoised d_{6}-d_{7} or d_{7} detail coefficients for segments with heart rate =93 bpm and <93 bpm, respectively. Six features characterizing the amplitude and area of the ICC and its first derivative (dICC) were calculated. Their PEA/PR discrimination power was measured using the area under the curve (AUC). These AUCs were compared with those obtained for the same features derived from the ICC/dICC extracted using an adaptive recursive least-squares (RLS) algorithm. The six features showed a mean (standard deviation) AUC of 0.91 (0.03) while RLS-based features yielded an AUC of 0.85 (0.07). Combining these ICC/dICC features with ECG features in a machine learning classifier might result in a robust pulse detector.
AB - An automated pulse detector during out-of-hospital cardiac arrest (OHCA) is needed. The thoracic impedance (TI) recorded through defibrillation pads presents an impedance circulation component (ICC), hidden among other components, in the form of small fluctuations correlated with each effective heartbeat. This study pre-sentes a method based on the stationary wavelet transform (SWT) to derive the ICC. A dataset with 456 5-s segments, 175 pulseless electrical activity (PEA) and 281 pulse-generating rhythm (PR), with concurrent ECG and TI signals from 49 OHCA patients was used. The SWT was used to decompose the TI into 7 levels. The ICC was derived from soft denoised d_{6}-d_{7} or d_{7} detail coefficients for segments with heart rate =93 bpm and <93 bpm, respectively. Six features characterizing the amplitude and area of the ICC and its first derivative (dICC) were calculated. Their PEA/PR discrimination power was measured using the area under the curve (AUC). These AUCs were compared with those obtained for the same features derived from the ICC/dICC extracted using an adaptive recursive least-squares (RLS) algorithm. The six features showed a mean (standard deviation) AUC of 0.91 (0.03) while RLS-based features yielded an AUC of 0.85 (0.07). Combining these ICC/dICC features with ECG features in a machine learning classifier might result in a robust pulse detector.
UR - http://www.scopus.com/inward/record.url?scp=85100928647&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85100928647&partnerID=8YFLogxK
U2 - 10.22489/CinC.2020.364
DO - 10.22489/CinC.2020.364
M3 - Conference contribution
AN - SCOPUS:85100928647
T3 - Computing in Cardiology
BT - 2020 Computing in Cardiology, CinC 2020
PB - IEEE Computer Society
T2 - 2020 Computing in Cardiology, CinC 2020
Y2 - 13 September 2020 through 16 September 2020
ER -