The retrospective study from December 2017 to June 2023, analyzed medical data from 173 COPD patients and 176 healthy controls. By applying deep learning segmentation modules, the team was able to automatically extract imaging features from different lung regions, including the parenchyma, airway, pulmonary arteries, and veins.
To refine accuracy, statistical methods such as the Mann–Whitney U-test and the least absolute shrinkage and selection operator (LASSO) were applied to identify the most relevant imaging markers. The machine learning models were trained using a support vector machine (SVM) classifier, tested internally, and further validated with an external dataset of 68 individuals.
For COPD diagnosis, the model achieved an area under the curve (AUC) of 0.981 in the training set and 0.977 in the testing set, both strong indicators of diagnostic reliability. Corresponding accuracy rates were 94.9% and 95.6%, respectively, this suggests that the system could distinguish COPD patients from healthy individuals with exceptional precision.
When applied to severity grading, the model demonstrated slightly lower but still clinically useful performance. The training set produced an AUC of 0.889 with an accuracy of 78.4%, while the testing set achieved an AUC of 0.796 and an accuracy of 71.9%. Although grading accuracy did not reach the same levels as diagnostic performance, this research noted that the results remain valuable for clinical application, especially when traditional lung function tests are impractical.
The findings highlight that automated imaging analysis could serve as a powerful adjunct to conventional COPD diagnosis and monitoring. By using chest CT scans, the AI model can predict lung function parameters and provide a reliable assessment of disease severity without invasive or strenuous testing.
Overall, these findings emphasize that while pulmonary function tests remain the gold standard, this approach has the potential to improve efficiency and accessibility, particularly in acute or resource-limited settings. The research suggested that machine learning–based CT analysis may soon become a critical tool in assisting physicians with earlier diagnosis and more tailored treatment plans for COPD patients.
Source:
Sui, H., Mo, Z., Wei, Y., Shi, F., Cheng, K., & Liu, L. (2025). Diagnosis and severity assessment of COPD based on machine learning of chest CT images. International Journal of Chronic Obstructive Pulmonary Disease, 20, 2853–2867. https://doi.org/10.2147/COPD.S528988
Disclaimer: This website is primarily for healthcare professionals. The content here does not replace medical advice and should not be used as medical, diagnostic, endorsement, treatment, or prescription advice. Medical science evolves rapidly, and we strive to keep our information current. If you find any discrepancies, please contact us at corrections@medicaldialogues.in. Read our Correction Policy here. Nothing here should be used as a substitute for medical advice, diagnosis, or treatment. We do not endorse any healthcare advice that contradicts a physician's guidance. Use of this site is subject to our Terms of Use, Privacy Policy, and Advertisement Policy. For more details, read our Full Disclaimer here.
NOTE: Join us in combating medical misinformation. If you encounter a questionable health, medical, or medical education claim, email us at factcheck@medicaldialogues.in for evaluation.