Objectives: The aim of this study is to compare the accuracy of the standard visual scoring of computed tomography (CT) scans with a cloud-based quantitative CT analysis that uses the StratX software, to measure collateral ventilation and, thus, predict lobar atelectasis after valve treatment. Methods: This is a retrospective, multicentre study of patients who had previously undergone valve treatment for severe heterogeneous emphysema and whose required fissure integrity ≥90% had been qualitatively scored by visual assessment of CT scans. For this study, all preprocedural CT scans were retrospectively analysed using the StratX software to provide quantitative scores of fissure integrity. The diagnostic accuracies of the visual and quantitative scores for predicting a target lobe volume reduction (TLVR) of ≥350 ml were calculated and statistically compared, as this level of volume reduction can be achieved only with sound fissure integrity. The clinical outcome of TLVR was also evaluated according to the minimal clinically important difference criteria. Results: Eighty-three patients were included in the analysis. Of them, 65 of 83 (78%) patients presented with TLVR ≥350 ml. Visual scoring correctly identified the absence of collateral ventilation in 65 of 83 (78%) cases but failed in 18 of 83 (22%) cases. Of these 18 patients, quantitative analysis showed that 16 of 18 (89%) patients did not present completeness of the fissure. The diagnostic accuracy of the quantitative analysis was better than that of the visual analysis (96.4% vs 78.3%; P = 0.0003). Only patients having TLVR ≥350 ml met or exceeded the minimal clinically important difference criteria. Conclusions: The quantitative analysis using the StratX software contributed a more objective and efficient evaluation of collateral ventilation that would have improved the selection of emphysematous patients for endobronchial valve treatment in the study population.
Visual analysis versus quantitative CT analysis of interlobar fissure integrity in selecting emphysematous patients for endobronchial valve treatment
Fiorelli A.;Cascone R.;Santini M.;
2019
Abstract
Objectives: The aim of this study is to compare the accuracy of the standard visual scoring of computed tomography (CT) scans with a cloud-based quantitative CT analysis that uses the StratX software, to measure collateral ventilation and, thus, predict lobar atelectasis after valve treatment. Methods: This is a retrospective, multicentre study of patients who had previously undergone valve treatment for severe heterogeneous emphysema and whose required fissure integrity ≥90% had been qualitatively scored by visual assessment of CT scans. For this study, all preprocedural CT scans were retrospectively analysed using the StratX software to provide quantitative scores of fissure integrity. The diagnostic accuracies of the visual and quantitative scores for predicting a target lobe volume reduction (TLVR) of ≥350 ml were calculated and statistically compared, as this level of volume reduction can be achieved only with sound fissure integrity. The clinical outcome of TLVR was also evaluated according to the minimal clinically important difference criteria. Results: Eighty-three patients were included in the analysis. Of them, 65 of 83 (78%) patients presented with TLVR ≥350 ml. Visual scoring correctly identified the absence of collateral ventilation in 65 of 83 (78%) cases but failed in 18 of 83 (22%) cases. Of these 18 patients, quantitative analysis showed that 16 of 18 (89%) patients did not present completeness of the fissure. The diagnostic accuracy of the quantitative analysis was better than that of the visual analysis (96.4% vs 78.3%; P = 0.0003). Only patients having TLVR ≥350 ml met or exceeded the minimal clinically important difference criteria. Conclusions: The quantitative analysis using the StratX software contributed a more objective and efficient evaluation of collateral ventilation that would have improved the selection of emphysematous patients for endobronchial valve treatment in the study population.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
