Background: e-Lung (Brainomix) is an artificial intelligence (AI)-driven software that is based on multi-class convolutional neural network (CNN) techniques. The aim of this research was to demonstrate the feasibility of e-Lung to evaluate progression in lung volume reduction in patients with interstitial lung disease (ILD) undergoing lung transplant assessments.
Methods: This was a single-center retrospective cohort study of consecutive patients with ILD who received lung transplants between June 2021 and November 2024. Patients who underwent serial prospective conventional evaluations using lung function testing (LFT) and conventional radiological assessments as well as retrospective lung volume measurements using e-Lung were included in this study.
Results: An analysis of 20 consecutive patients who met strict inclusion criteria and underwent an additional e-Lung assessment revealed that both the serial physiological actual total lung capacity (aTLC) measurements and e-Lung-derived lung volume measurements were able to provide recipient lung size estimations and detect serial declines in lung volume. A poorer DLCO (2.61 ± 0.77 vs. 3.87 ± 1.59 mmol/min/kPa, p = 0.044) at the time of wait-listing was associated with a significant lung volume reduction.
Conclusions: e-Lung may serve as an additional upscale tool for the rapid and objective quantitative evaluation of the actual lung volume and the detection of the extent of parenchymal shrinking in patients with advanced ILD awaiting lung transplantation.
