Automated classifier detected lung fibrosis in scleroderma patients
Click Here to Manage Email Alerts
A computer-aided diagnosis system is able to detect lung fibrosis in patients with scleroderma, according to study results.
Researchers used this computer-aided diagnosis (CAD) system to detect the presence or absence of pulmonary fibrosis in patients with scleroderma-related interstitial lung disease (ILD) within 1% and a clinically meaningful threshold of 25% using high-resolution CT images.
High-resolution, thin-section CT images were obtained and analyzed from the record of 129 adult patients (36 men) who were part of the Scleroderma Lung Study.
Researchers used the following criteria to create a qualified lung fibrosis score:
• images were denoised;
• images were grid sampled;
• the characteristics of grid intensities were converted into texture features;
• texture features classified pixels as fibrotic or non-fibrotic, with fibrosis defined by a reticular; pattern with architectural distortion; and
• fibrotic pixels were reported as percentages.
“Our findings suggest that this automated classifier is potentially useful for reproducible objective measurements of fibrosis in clinical trials of interventions in ILD,” researchers concluded.