CT evaluation of interstitial lung disease related to systemic sclerosis: visual versus automated assessment. A systematic review.

AIM: To identify similarities and differences between visual (VA) and automated assessment (AA) of systemic sclerosis-related interstitial lung disease (SSc-ILD) at chest computed tomography (CT) in terms of clinical applicability. MATERIALS AND METHODS: Medline, Embase, and Web of Science were searched to identify all studies investigating VA and AA for SSc-ILD assessment, from inception to 31 July 2022. Exclusion criteria were manuscripts not in English, absence of full-text, reviews, diseases other than ILD in SSc, CT not analysed with both VA and AA, VA and AA not adopted for the same purpose or not compared, overlap syndromes, SSc-ILD data not extractable, and studies with <10 patients. RESULTS: Ten full-text studies (804 patients) were included. The most adopted VAs were the Warrick or Goh score (four studies each), while densitometry (eight studies) or lung texture analysis (LTA, two studies) were utilised as AAs. The main field of investigation was the correlation with baseline pulmonary function tests (PFT, six studies). Warrick VA showed lower correlations compared to densitometry, while Goh VA demonstrated more heterogeneous results. Compared to LTA, Goh VA obtained lower correlations with lung volumes but similar or stronger coefficients with alveolar diffusibility. CONCLUSIONS: VA and AA may show heterogeneous results comparing their correlations with PFT, probably depending on the specific analysis adopted for each method. More data are needed on VA versus LTA. Comparisons between VA and AA regarding correlation with PFT follow-up and as prognostic elements, or for disease monitoring, are lacking. AAs in progressive fibrosis diagnosis remain to be tested.

Computed tomography - histology correlations of unusual lung tumors.

A large variety of rare benign and malignant tumors may sporadically affect the lung. Computed tomography (CT) findings of unusual primary lung tumors are often nonspecific. However, there are some rare pulmonary tumors with imaging features overlapping those of other conditions, thus making radiologic diagnosis challenging. The aim of this review was to correlate CT and histopathological features of a variety of unusual lung tumors to better clarify when and to what extent radiological diagnosis is reliable.