Deep Learning for The Chronic Obstructive Pulmonary Disease Assessment using Lung CT

Authors

  • Halimah Alsurayhi Westren University
  • Abbas Samani Department of Electrical and Computer Engineering, Western University, London, ON, Canada 2 Biomedical Engineering School, Western University, London, ON, Canada 3 Departments of Medical Biophysics, Western University, London, ON, Canada

Abstract

COPD is a prevalent and progressive lung disease. It is identified as airflow limitation due to airway and/or air sacs inflammation. To limit disease progression and prevent future exacerbation, GOLD has recently proposed a staging system to devise proper treatment plan. Predictors used in this staging system assessed using a comprehensive or simple questionnaire about the disease history. In contrast to such qualitative assessment approach, lung CT images can be used for quantitative assessment of the disease severity.  In this project, we developed a classification tool for COPD assessment using lung CT images processed through a deep learning model. For this purpose, we developed two CNN models to identify the severity of the two predictors of symptoms and exacerbation. The two models are utilized to design a novel classification tool based on the GOLD2023 staging system. Results obtained from this model indicate reasonable accuracy despite requiring a single 3D CT scan.

Downloads

Published

2024-06-26

How to Cite

[1]
H. Alsurayhi and A. Samani, “Deep Learning for The Chronic Obstructive Pulmonary Disease Assessment using Lung CT ”, CMBES Proc., vol. 46, Jun. 2024.

Issue

Vol. 46 (2024)

Section

Academic