Do we need computational analysis of high-resolution CT images in fibrotic interstitial lung disease?

Computational analysis is showing that the diagnostic and prognostic role of radiological images goes much beyond what can be captured by human eyes. High-resolution computed tomography (HRCT) has an established role in the management of fibrotic interstitial lung disease (f-ILD) despite some major limits. Visual analysis of HRCT images is affected by high intra- and interobserver variability. Moreover, visual analysis has limits as a prognostic biomarker in f-ILD patients; it is well known that a usual interstitial pneumonia (UIP) pattern on HRCT is associated with a poor outcome, but this distinction is imprecise since also patients with a UIP pattern can have a variable disease course. Computational analysis of HRCT images has the potential to overcome these limits having already outperformed radiologists for quantification of disease extent at a single time-point, detection of changes during follow-up and as a prognostic biomarker.