Radiological Insights into UIP Pattern: A Comparison Between IPF and Non-IPF Patients

Abstract

Background/Objectives: This study aims to distinguish radiological differences between primary idiopathic Usual Interstitial Pneumonia (UIP) and secondary UIP patterns Methods: This retrospective study included patients with HRCT findings consistent with a UIP pattern. Final diagnoses were established via multidisciplinary discussion and classified as primary UIP/IPF or secondary UIP, following the 2022 ATS/ERS/JRS/ALAT guidelines. An expert thoracic radiologist (>10 years of experience), blinded to clinical data, reviewed the earliest available HRCT assessing key imaging features: honeycombing (micro-, macro- or exuberant), fibrosis distribution (symmetry, anterior-upper lobe sign, etc.), ground-glass opacities (GGO), dilatation of esophagus. Additionally, AI software AVIEW Build 1.1.46.28-win Coreline (©Coreline Soft Co., Ltd. All Rights Reserved). performed lung texture analysis, quantifying total lung volume and radiological patterns. Statistical analysis was performed to reveal results. Results: Among 53 cases, 31 were classified as IPF and 22 as secondary UIP cases. The expert radiologist achieved a diagnostic sensitivity of 82.9%, specificity of 889%, with a positive predictive value of 93.5%-in distinguishing between primary and secondary UIP. Primary UIP cases exhibited typical hallmark radiological features, including uniform honeycombing with cranio-caudal distribution (90.3%). Reticulations contributed significantly to the fibrotic texture, maintaining a consistent cranio-caudal gradient and axial symmetry (84.8%). Secondary UIP displayed more significant radiological heterogeneity, including patchy fibrosis with irregular GGO distribution (84.5% versus 53.33%); other findings-such as exuberant honeycombing, four corner sign and wedge-shaped fibrosis-were mainly observed in secondary pattern with respective percentages of 31.8%, 9% and 49%. Conclusions: Experienced thoracic radiologists, leveraging hallmark imaging features, play a critical role in improving diagnostic accuracy between primary and secondary UIP patterns.

Keywords: AI-software; HRCT; UIP.

Figure 1

High-resolution computed tomography (HRCT) of the chest in axial (a,b) views. (a) Axial HRCT shows ground-glass opacities (subtle hazy increased attenuation) in left upper lobe, superimposed on fine reticulations (red arrow), suggesting fibrotic interstitial lung disease. (b) Advanced fibrotic changes are evident with traction bronchiectasis (yellow arrowhead) and extensive honeycombing (white arrow) in the posterior basal segments of the lower lobes, characterized by clustered cystic airspaces of similar diameter.

Figure 2

High-resolution computed tomography (HRCT) of the chest in axial (left) and coronal (right) views showing exuberant honeycombing, characterized by extensive, confluent subpleural cystic airspaces with well-defined walls and similar diameters, predominantly in the lower lobes and posterior regions.

Figure 3

High-resolution computed tomography (HRCT), axial images (a,b). Predominantly distribution in the ventral-anterior regions of both upper lobes—as clearly demonstrated on Figure 3a,b (white circles).

Figure 4

Coronal HRCT images of the chest showing the “straight edge sign” (white arrows in (a,b), characterized by sharp demarcation of fibrosis with a straight, horizontal lower boundary, limited to the lung bases without significant extension along the lateral margins. This imaging feature is often associated with connective tissue disease-related interstitial lung disease (CTD-ILD) and helps differentiate it from idiopathic pulmonary fibrosis (IPF).

Figure 5

Box whisker plots representing comparisons between primary UIP and secondary UIP—for honeycombing (HC), reticulations (RET), ground-glass (GG), consolidations (Cons), enphysema (E) and Volume (Vol).