Abstract
Current methods for quantifying perfusion from computed tomography pulmonary angiography (CTPA) often rely on semi-quantitative scoring systems and requires an experienced evaluator. Few studies report on absolute quantitative variables derived from the images, and the methods are varied with mixed results. Dual-energy CTPA (DE-CTPA) enables automatic quantification of lung and lobar perfusion with minimal user interaction by utilizing machine learning based software. We aimed to evaluate differences in DE-CTPA derived quantitative perfusion variables between patients with acute pulmonary embolism (PE) and chronic thromboembolic pulmonary hypertension (CTEPH). This retrospective, single-center, observational study included 162 adult patients diagnosed with PE (n = 81) or CTEPH (n = 81) and scanned using dual-energy CT between 2020 and 2023. Mann-Whitney U tests and permutational analysis of variance (PERMANOVA) were used for comparative analyses. We found whole lung perfusion blood volume to be lower (p < 0.001) in PE patients (median 3399 mL [2554, 4284]) than in CTEPH patients (median 4094 mL [3397, 4818]). The same was observed at single lung and lobar level. PERMANOVA encompassing all perfusion variables showed a difference between the two groups (F-statistic = 13.3, p = 0.002). Utilizing logistic regression, right and left lower lobe perfusion blood volume showed some ability to differentiate between PE and CTEPH with area under the receiver operation characteristics curve values of 0.71 (95% CI: 0.56; 0.84) and 0.72 (95% CI: 0.56; 0.86). Pulmonary perfusion is lower in patients with PE than patients with CTEPH, highlighted by differences in DECT-derived perfusion blood volume. Quantitative perfusion variables might be useful to differentiate between the two diseases.