TY - GEN
T1 - Identifying Optimal Features from Heart Rate Variability for Early Detection of Sepsis in Pediatric Intensive Care
AU - Amiri, Paria
AU - Derakhshan, Amin
AU - Gharib, Behdad
AU - Liu, Ying Hsang
AU - Mirzaaghayan, Mohamadreza
N1 - Publisher Copyright:
© 2019 IEEE.
PY - 2019/7
Y1 - 2019/7
N2 - Sepsis as bacterial infection is the most common and costly causes of mortality in critically ill patients. The early diagnosis of sepsis is significantly important for effective treatment. In this study, over a period of two years, the electrocardiogram of nearly 500 pediatric and neonate patients with heart diseases were collected in 24 hours before diagnosis. The collected data of 22 patients were studied including 11 sepsis patients with positive blood cultures and 11 non-sepsis patients. After extracting the HRV (Heart Rate Variability) signal, 28 linear and nonlinear features according to previous research were extracted. By using the relative entropy method as a feature selection technique, the extracted features were evaluated for their ability to discriminate the data in sepsis and non-sepsis groups, and the best features were entered into the classification process. Using the four classification models of SVM, LDA, KNN and Decision Tree, the accuracy of 86.36% was obtained with Decision Tree for discrimination of sepsis patients from other patients.
AB - Sepsis as bacterial infection is the most common and costly causes of mortality in critically ill patients. The early diagnosis of sepsis is significantly important for effective treatment. In this study, over a period of two years, the electrocardiogram of nearly 500 pediatric and neonate patients with heart diseases were collected in 24 hours before diagnosis. The collected data of 22 patients were studied including 11 sepsis patients with positive blood cultures and 11 non-sepsis patients. After extracting the HRV (Heart Rate Variability) signal, 28 linear and nonlinear features according to previous research were extracted. By using the relative entropy method as a feature selection technique, the extracted features were evaluated for their ability to discriminate the data in sepsis and non-sepsis groups, and the best features were entered into the classification process. Using the four classification models of SVM, LDA, KNN and Decision Tree, the accuracy of 86.36% was obtained with Decision Tree for discrimination of sepsis patients from other patients.
UR - http://www.scopus.com/inward/record.url?scp=85077880120&partnerID=8YFLogxK
U2 - 10.1109/EMBC.2019.8856346
DO - 10.1109/EMBC.2019.8856346
M3 - Conference contribution
T3 - Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS
SP - 1425
EP - 1428
BT - 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2019
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2019
Y2 - 23 July 2019 through 27 July 2019
ER -