Skip to Main Content
Frequently Asked Questions
Submit an ETD
Global Search Box
Need Help?
Keyword Search
Participating Institutions
Advanced Search
School Logo
Files
File List
MULTI-PARAMETER PHYSIOLOGICAL TRACKING SYSTEM FOR DIAGNOSIS OF SEPSIS_222021.pdf (4.74 MB)
ETD Abstract Container
Abstract Header
MULTI-PARAMETER PHYSIOLOGICAL TRACKING SYSTEM FOR DIAGNOSIS OF SEPSIS
Author Info
Guo, Fei
Permalink:
http://rave.ohiolink.edu/etdc/view?acc_num=case1613129350013438
Abstract Details
Year and Degree
2021, Doctor of Philosophy, Case Western Reserve University, EECS - Electrical Engineering.
Abstract
Sepsis is a systemic inflammatory response to infection that can progress to septic shock and multi-organ dysfunction and a life-threating situation without proper and prompt clinical treatment. It causes numerous alterations to the human dynamic system that if quantified can provide diagnostic and prognostic insights. To address this issue, this dissertation proposes a multi-channel physiological signal analysis system that elucidates the association of characteristic alterations of the cardiovascular and ventilatory systems along with their impact on sepsis. Variability analysis techniques depict biological time series with regards to the fluctuations, spectral composition, scale-free variations and degrees of regularity or complexity. Specifically, multi-dimensional noninvasive biomarkers are generated from Heart Rate Variability (HRV) analysis, Respiratory Rate Variability (RRV) analysis and Blood Pressure Variability (BPV) analysis. In this dissertation, linear and nonlinear analysis including Time-Frequency domain analysis, Detrended Fluctuation Analysis (DFA), Multiscale Entropy (MSE) and Poincare analysis can differentiate pathological sepsis patients from normal ICU patients with statistically significant levels. In addition, a septic prediction framework using low-density vital signs is proposed to provide early warning indicators of the onset of sepsis for clinical personnel. The proposed early prediction index generated from LSTM network and the ensemble XGBoosting classifier overcame challenges from a high percentage of unavailable data and extreme unbalanced classes challenges to reach prediction results with the AUROC at 0.8132 and 0.8255, respectively.
Committee
Kenneth Loparo (Committee Chair)
Farhad Kaffashi (Committee Member)
Thomas Dick (Committee Member)
Frank Jacono (Committee Member)
Pages
118 p.
Subject Headings
Biomedical Research
;
Electrical Engineering
Recommended Citations
Refworks
EndNote
RIS
Mendeley
Citations
Guo, F. (2021).
MULTI-PARAMETER PHYSIOLOGICAL TRACKING SYSTEM FOR DIAGNOSIS OF SEPSIS
[Doctoral dissertation, Case Western Reserve University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=case1613129350013438
APA Style (7th edition)
Guo, Fei.
MULTI-PARAMETER PHYSIOLOGICAL TRACKING SYSTEM FOR DIAGNOSIS OF SEPSIS.
2021. Case Western Reserve University, Doctoral dissertation.
OhioLINK Electronic Theses and Dissertations Center
, http://rave.ohiolink.edu/etdc/view?acc_num=case1613129350013438.
MLA Style (8th edition)
Guo, Fei. "MULTI-PARAMETER PHYSIOLOGICAL TRACKING SYSTEM FOR DIAGNOSIS OF SEPSIS." Doctoral dissertation, Case Western Reserve University, 2021. http://rave.ohiolink.edu/etdc/view?acc_num=case1613129350013438
Chicago Manual of Style (17th edition)
Abstract Footer
Document number:
case1613129350013438
Download Count:
47
Copyright Info
© 2021, all rights reserved.
This open access ETD is published by Case Western Reserve University School of Graduate Studies and OhioLINK.