MSc Defence: Principal Component Analysis of Quantitative Computed Tomography Features and Visual Emphysema Scores: Association with Lung Function Decline

Student: Meghan Koo

Supervisor: Dr. Miranda Kirby

Abstract:

Emphysema is characterized by the damage of the alveoli (air sacs) in the lungs and occurs in
patients with Chronic Obstructive Pulmonary Disease (COPD). It is clinically evaluated visually
by a radiologist who subjectively scores the severity of the disease on computed tomography (CT)
images. However, there are limitations to this method, including inherent inter- and intra-observer
variability. A solution to overcome these limitations is the use of fully automated and objective
quantitative CT (QCT) measurements. However, it has been shown in literature that visual score
outperforms single QCT measurements when predicting mortality, despite its subjectivity. In this
thesis, we explore the use of multiple QCTs in combination by principal component analysis
(PCA), and its ability to predict COPD disease progression (lung function decline), independent
of visual scoring.