Introduction
Emphysema severity, emphysema heterogeneity, and pulmonary fissure completeness are predictors of response to endobronchial valve reduction treatment.1 In particular, the assessment of pulmonary fissure completeness is of
paramount importance due to being a reliable indicator of collateral ventilation.2,3 Fissure completeness and emphysema severity can be determined visually, which is time-consuming and suffers from inter- and intra-observer variability.4,5
Traditional computer algorithms designed for the quantitative assessment of fissure completeness and emphysema severity are based on classical pattern recognition techniques and have provided good results in the past.6,7 With the
advent of deep learning-based artificial intelligence (AI) techniques, however, newer algorithms now hold promise to provide a more accurate and reliable approach to the assessment of fissure completeness and emphysema.
In general, deep learning algorithms improve their performance when provided with more data and therefore it is expected that fissure completeness and emphysema severity assessment performance will increase over time.
The purpose of this study was to test the performance of the SeleCT service with Imbio, a novel AI based software package for the assessment of fissure completeness, emphysema severity, and emphysema heterogeneity in a cohort of clinical patients with emphysema deemed to be candidates for endobronchial valve placement.