Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article

Degree

Doctor of Philosophy

Program

Electrical and Computer Engineering

Supervisor

Abbas Samani

Abstract

Chronic Obstructive Pulmonary Disease (COPD) is a condition characterized by persistent inflammation and airflow blockages in the lungs, contributing to a significant number of deaths globally each year. To guide tailored treatment strategies and mitigate future risks, the Global Initiative for Chronic Obstructive Lung Disease (GOLD) employs a multifaceted assessment system of COPD severity, considering patient's lung function, symptoms, and exacerbation history. COPD staging systems, such as the high-resolution eight-stage COPD system and the GOLD 2023 three staging systems, have been later developed based on these factors. Lung Computed Tomography (CT) is becoming increasingly crucial in investigating COPD as it can detect various COPD phenotypes, such as emphysema, bronchial wall thickening, and gas trapping. Deep learning techniques show promise in leveraging CT imaging to assess the severity of COPD. This thesis uses lung CT data in conjunction with machine learning techniques to classify COPD patients according to these staging systems. For the eight-stage system, both Neural Network and Convolutional Neural Network (CNN) approaches were employed for classification. To develop the Neural Network model, features were extracted from lung CT scans at inspiration and expiration breathing phases, including lung air features and COPD phenotypes features. The CNN model utilized a single lung CT scan at the expiration phase. The GOLD 2023 three staging system involves training separate CNN models using lung CT scans at expiration to predict symptom levels and COPD exacerbation risk. In this thesis, in addition to models trained from scratch, Transfer Learning was also employed to develop models for the eight-stage COPD classification, Symptom level prediction, and exacerbation risk prediction. The developed classifiers demonstrate reasonably high classification performance, indicating their potential for deployment in clinical settings to enhance COPD assessment using image data.

Summary for Lay Audience

Summary Chronic Obstructive Pulmonary Disease (COPD) is a health condition that affects the lungs and can cause inflammation and blockages in their airways. It is a leading cause of death worldwide. To help treat COPD, the Global Initiative for Chronic Obstructive Lung Disease (GOLD) uses three factors to assess the disease's severity: the patient's lung function, symptoms, and history of exacerbation (flare-ups). There are two classifications used to determine the stage of COPD: the eight-stage system and the GOLD 2023 three staging systems. These staging systems use the three factors to determine the severity of the disease. Lung computed tomography (CT) images can be used to investigate COPD because they can detect different types of conditions, including lung tissue destruction, airway wall thickening, and air trapping in the lungs. Machine learning techniques can be used to assess the severity of COPD automatically using CT imaging. In this study, machine learning was used to classify COPD patients according to the eight-stage and GOLD 2023 three staging systems. Two machine learning methods were used to develop the eight-stage classifiers: Neural Networks (NN) and Convolutional Neural networks (CNN). For the NN model, lung imaging algorithms were used to extract features from paired lung CT at the inhalation and exhalation breathing phases. To develop the GOLD2023 classifier, two separate CNN models were also trained to predict symptom levels and the risk of COPD exacerbation. The developed classifiers showed promising results, demonstrating their potential for clinical use in improving COPD assessment using image data.

Download

Share

COinS