Evolving Role of Quantitative CT in Interstitial Lung Disease from Technical Challenges to Clinical Implementation

Abstract

Interstitial lung diseases (ILDs) comprise a heterogeneous group of more than 200 disorders affecting the lung parenchyma, marked by varying degrees of inflammation and fibrosis. High-resolution computed tomography (HRCT) remains central to ILD diagnosis and classification, yet visual interpretation is limited by subjectivity and substantial interobserver variability, reducing reliability for baseline assessment and longitudinal follow-up. Quantitative CT (qCT) has emerged as a powerful solution, offering objective, reproducible measurements of lung density, texture, vascular remodelling, and fibrosis extent. This systematic review evaluates the evolving role of qCT in ILD across diagnostic assessment, prognostication, disease monitoring, and emerging machine learning based applications. Following PRISMA guidelines, a comprehensive literature search identified 185 eligible studies. Across studies, qCT-derived biomarkers showed strong correlations with physiologic impairment, including forced vital capacity and diffusing capacity, and reliably predicted disease progression and mortality in idiopathic pulmonary fibrosis, connective tissue disease associated ILD, and hypersensitivity pneumonitis. Machine learning and deep learning approaches further improved segmentation accuracy, pattern recognition, and prediction of clinically meaningful outcomes, expanding the potential of qCT as a sensitive imaging biomarker. Despite these advances, several barriers limit routine clinical adoption. Variability in acquisition protocols, reconstruction methods, segmentation techniques, and feature extraction reduces reproducibility across scanners and institutions. Standardization efforts, combined with robust external validation and integration of explainable AI, are crucial for translating quantitative tools into practice. Overall, qCT represents a significant advancement in ILD imaging, with the potential to enhance diagnostic confidence, improve risk stratification, and enable more precise monitoring of therapeutic response.