Interstitial lung abnormalities (ILA) on routine chest CT: Comparison of radiologists’ visual evaluation and automated quantification

•Interreader agreement for presence of ILA based on the Fleischner society guideline was substantial (κ, 0.74).•Interreader agreement for the subtype of ILA was only fair (κ, 0.38).•The deep learning-based quantitative system for ILA was able to identify ILA despite the use of various CT protocols.•Automated quantification tool dedicated for ILA in routine clinical practice may aid in objectively identifying ILA by accurately quantifying the extent of ILA per zonal volume.