High-Resolution Computed Tomography combined with clinical information may help detect COPD early: Study

This study was published in the International Journal Of Chronic Obstructive Pulmonary Disease by Zhang W. and colleagues. This study aimed to identify high-risk COPD participants by employing QCT measurements in conjunction with pulmonary function tests (PFTs) and clinical characteristics. 140 participants diagnosed as Non-COPD (FEV1/FVC > 70%) were divided into low-risk COPD (FEV1% ≥ 95%) and high-risk groups (80% < FEV1% < 95%).

Inspiratory high-resolution CT scans were conducted along with PFTs and clinical data collection. QCT measurements were analyzed using statistical methods like Student’s t-test, Mann–Whitney U-test, LASSO method for feature identification, and Spearman rank correlation test for assessing associations with PFTs. Support Vector Machine (SVM) was employed to identify high-risk COPD participants, evaluating the model’s performance in terms of accuracy, sensitivity, specificity, F1-score, and AUC.

• SVM Performance: The SVM model utilizing QCT measurements exhibited promising accuracy in identifying high-risk COPD patients with 85.71% accuracy, 88.34% sensitivity, 84.00% specificity, 83.33% F1-score, and an AUC of 0.93.

• Integration with Clinical Data: Incorporating QCT measurements with clinical information further improved the model's performance, achieving an accuracy of 90.48%.

• Correlation with PFTs: Certain QCT measurements, particularly the emphysema index of the left lower lung (%LAA−950), showed a negative correlation with PFTs, signifying their relevance in assessing pulmonary function status.

The integration of QCT indexes into clinical practice could significantly aid in evaluating pulmonary function status, potentially allowing for early interventions and improved management strategies for COPD patients.

Reference: