A machine learning model based on chest CT images accurately diagnoses and grades the severity of COPD.
Key Details
- 1Researchers developed a machine learning model using chest CT data from 173 COPD patients and 176 healthy controls.
- 2The model segments the lung parenchyma, airway, pulmonary artery, and vein, then extracts imaging features.
- 3Diagnostic accuracy for COPD was 95% (training set) and 96% (test set); AUC was 0.98 and 0.97, respectively.
- 4Severity grading accuracy was 78% (training) and 72% (test); AUC was 0.89 and 0.8.
- 5Traditional spirometry tests may be less effective or difficult for some patients, highlighting CT's added value.
Why It Matters
This machine learning approach could offer a more accessible and detailed assessment for COPD diagnosis and severity, especially for patients who struggle with spirometry. Improved diagnostic performance has the potential to personalize care and improve clinical outcomes for COPD patients in radiology practice.
Source
AuntMinnie
Related News
•Radiology Business
AI Triage Cuts CT Report Turnaround for Pulmonary Embolism—Daytime Only
FDA-backed study finds AI triage tools reduce radiology CT report turnaround times for pulmonary embolism during peak hours.
•Cardiovascular Business
AI Uses Mammograms to Predict Women’s Cardiovascular Disease Risk
AI algorithms can analyze mammograms to predict cardiovascular disease risk, expanding the utility of breast imaging.
•Health Imaging
BMI Significantly Impacts AI Accuracy in CT Lung Nodule Detection
New research demonstrates that high BMI negatively impacts both human and AI performance in chest low-dose CT interpretation, highlighting dataset diversity concerns.