Deep Learning-Based Estimated Pulmonary Biological Age From Chest Computed Tomography Images in Healthy Adults: Model Development and Validation Study.

Deep learning models demonstrated acceptable applicability for estimating pulmonary biological age from chest CT and showed strong correlation between ePBA and chronological age.

• Multiple deep learning models were trained and evaluated for this task.
• Training data comprised 7,726 chest CT scans from institution A.
• External validation was performed on CT scans from institution B (n=1,506) and institution C (n=1,955).
• Total dataset included 11,187 chest CT scans from healthy adults across 3 health management centers.

Age gap (ePBA minus chronological age) was significantly associated with forced expiratory volume in 1 second (FEV1) reduction in patients with COPD.

• The association was assessed among 138 patients with COPD hospitalized at institution A during the same time period.
• Age gap was negatively correlated with FEV1 expressed as percentage of predicted values (rs = -0.18; P = .03).
• This indicates that a larger age gap (biologically older lungs) was associated with worse pulmonary function.

Age gap was significantly associated with an increased risk of all-cause mortality in patients with COPD.

• The analysis was conducted in 138 hospitalized COPD patients from institution A.
• Each unit increase in age gap was associated with a hazard ratio of 1.16 (95% CI 1.08–1.25) for all-cause mortality.
• The finding suggests ePBA provides prognostic information beyond chronological age in COPD patients.

The study identified a need for large-scale healthy adult-based training data to improve generalizability of CT-based ePBA models.

• Prior models were noted to lack training and validation with large-scale healthy adults, hindering generalizability.
• This study addressed that gap by using 11,187 chest CT scans from healthy adults at 3 health management centers.
• Multicenter data collection was used to support external validation and broader applicability.

Estimated pulmonary biological age (ePBA) was proposed as a more reliable indicator for disease progression and mortality than chronological age.

• Chest CT was identified as a promising tool for calculating ePBA.
• The age gap metric (ePBA minus chronological age) was investigated as a novel clinical biomarker.
• Associations were examined for both pulmonary function outcomes and all-cause mortality in COPD patients.