Effort-Independent Asthma Severity Classification
Name
Lynch-jclynch-SM-EECS-2021-thesis.pdf
Description
Thesis PDF
Size
1.3 MB
Format
Adobe PDF
Checksum (MD5)
47104eac8ed3cae6ccad6ff5645cad7d
Author(s)
Lynch, James C., III
Advisor(s)
Heldt, Thomas
Date Issued
June 2021
Publisher
Massachusetts Institute of Technology
Abstract
Asthma is an obstructive pulmonary disorder. It impacts the lives of over 24 million individuals in the United States alone, a large segment of which are children. We propose to investigate capnography as a viable diagnostic modality to guide the treatment of asthma as an alternative to the gold standard, spirometry. Capnography shows promise in the detection of similar pulmonary disorders, and would serve as a noninvasive and effort-independent tool, providing critical information to clinicians when patients are unable or unwilling to comply with spirometry testing. In this work, we demonstrate the viability of using features extracted from time-based capnography to determine underlying patient symptom severity, using logistic regression classification models. Applications in both controlled, pulmonary function laboratories and emergency department triage conditions are explored. We show that for an adult population undergoing methacholine challenge pulmonary function testing, capnography recordings from subjects with asthmatic exacerbation may be distinguished from their normal/baseline recordings with an AUROC of 0.92 (0.84 -- 1.00). Additionally, using data from an acute pediatric setting we show that recordings from subjects with severe asthmatic exacerbation may be distinguished from subjects with mild or moderate asthma symptoms with an AUROC of 0.86 (0.72 -- 1.00).
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
In Copyright - Educational Use Permitted
Copyright MIT
Persistent DSpace Link
