Quantification differences between supine and prone CT in interstitial lung disease

Area of Science:

• Radiology and Imaging
• Pulmonary Medicine
• Medical Diagnostics

Background:

• Interstitial lung disease (ILD) assessment often relies on computed tomography (CT).
• Quantitative CT (QCT) offers objective measurements of lung abnormalities.
• Evaluating measurement variability between supine and prone imaging is crucial for QCT reproducibility.

Purpose of the Study:

• To compare measurement variability and reproducibility of prone QCT versus supine QCT in patients with ILD.
• To assess the correlation of QCT measurements with pulmonary function test (PFT) parameters in different positions.
• To identify factors influencing inter-positional variability in QCT measurements.

Main Methods:

• Retrospective analysis of paired supine and prone CT scans from 277 ILD patients (Jan 2021-Dec 2023).
• Lung abnormalities quantified using deep learning software; variability assessed by Bland-Altman and intraclass correlation coefficients (ICC).
• Correlations with PFTs evaluated using Pearson's correlation and Steiger's Z-test; regression analyses identified variability predictors.

Main Results:

• High reproducibility between supine and prone QCT (ICC > 0.75) with low variability (approx. 2% for fibrosis, 4% for total ILD extent).
• Ground-glass opacity (GGO) exhibited the highest variability, especially in extensive disease.
• Prone honeycombing showed stronger PFT correlation; hypersensitivity pneumonitis associated with greater total ILD variability.

Conclusions:

• Prone QCT demonstrates high reproducibility and comparable PFT correlations to supine QCT.
• Prone CT is reliable for single-time-point ILD assessments.
• Longitudinal comparisons require caution in advanced or GGO-dominant ILD due to potential variability.