CT-based radiomics features can help diagnose COPD earlier than ever before
Radiomics could play a valuable diagnostic role in the identification and severity staging of chronic obstructive pulmonary disease (COPD) when applied to chest CT images, new research suggests.
COPD has high morbidity and mortality rates and diminishes the quality of life in those who suffer from it. Early detection has proven to reduce mortality and slow progression, but pulmonary function tests (PFTs), which are the standard for COPD diagnosis, lack sensitivity in catching the disease in its preliminary stages.
“It is limited by insensitivity to early stages of COPD and lacks reproducibility,” corresponding author Kewu Huang, MD, with the department of pulmonary and critical care medicine at Beijing Chao-Yang Hospital, and co-authors explained. “Several studies suggest that when using spirometry alone, COPD is frequently either misdiagnosed or missed as a diagnosis entirely.”
Many characteristics of COPD—such as parenchymal destruction, bronchial wall thickening and interstitial lung abnormalities—can be identified on chest CT scans. This knowledge prompted researchers to inquire about the role radiomics could play for early diagnosis and severity staging of COPD using a radiomic signature based on chest CT images.
“Radiomic features can capture tissue and lesion characteristics such as heterogeneity,” the doctors noted.
For their research, the experts used the chest CT images of 322 participants (249 with COPD and 73 control patients) to extract 1,395 radiomic features. Two classification methods, support vector machine (SVM) and logistic regression (LR), were then used to identify and classify COPD and its severity.
The 1,395 radiomic features were narrowed down to 38 and 10 for identification and severity to create the signature. Using those features, both models (SVM and LR) performed well and had 90% or higher accuracy, sensitivity and specificity for identification and severity staging.
“Our study has shown that radiomics based applications offer the exciting prospect for identification and severity staging of COPD,” the doctors suggested. “We believe that this method is a powerful technique to identify patients who have not been previously identified as COPD in the general population.”
The doctors believe this method will lead to earlier diagnosis of COPD and could signal providers to begin appropriate management more promptly.
You can view the detailed research in Academic Radiology.
In addition to her background in journalism, Hannah also has patient-facing experience in clinical settings, having spent more than 12 years working as a registered rad tech. She began covering the medical imaging industry for Innovate Healthcare in 2021.