Deep-learning based quantitative evaluation of postoperative atelectasis following right upper lobectomy

Abstract Objective: Chronic postoperative atelectasis of lung lobes is an occasional complication following thoracic surgery, particularly middle lobe atelectasis following right upper lobectomy (RULobectomy). Existing methods of grading atelectasis are typically manual, subjective, and not scalable. We aimed to develop and validate an automated, deep learning–based volumetric framework to quantify and grade postoperative atelectasis using pre- and post-operative CT scans. Methods: We retrospectively included all patients who underwent RULobectomy in our institution from 2008 to 2023 who had available pre-operative and 6-month postoperative CT scans. We trained two separate nnU-Net v2 segmentation models for preoperative and postoperative lobar anatomy followed by volumetric quantification of the right middle lobe (RML), right lower lobe (RLL), and total lung volume. Atelectasis severity in the RML was independently graded by two surgeons using a standardized, 5-point radiological scale (none, minimal, subsegmental, segmental, lobar). The association between volume metrics and clinical atelectasis severity was evaluated using both the original 5-point scale and a pooled 3-point scale (none, minimal–subsegmental, segmental–lobar). Results: 236 patients comprised the study cohort. The pre- and postoperative models achieved high segmentation accuracy (mean Dice scores: 0.98 ± 0.02 and 0.99 ± 0.00, respectively). Median (IQR) RML volume loss progressively increased with higher atelectasis grades, from − 4.6 mL (-78.5, 59.0) in grade 0 to − 317.8 mL (-440.7, -194.8) in grade 4 atelectasis (p < 0.001). Normalized RML/right lung (RL) and RML/total lung (TL) volume ratios showed statistically significant differences across the pooled atelectasis grades (p < 0.001). Conversely, normalized RLL volumes increased with worsening RML atelectasis (p < 0.001), suggesting compensatory hyperinflation. Conclusions: We demonstrate the feasibility and clinical relevance of deep learning–based volumetric assessment of atelectasis after RULobectomy. This automated pipeline with its open-source model enables reproducible quantification of atelectasis through lobar volume loss and may serve as a scalable tool for clinical and research applications involving atelectasis.