Deep Learning Classification of Usual Interstitial Pneumonia Predicts Outcomes

Rationale: Computed tomography (CT) enables noninvasive diagnosis of usual interstitial pneumonia (UIP), but enhanced image analyses are needed to overcome the limitations of visual assessment. Objectives: Apply multiple instance learning (MIL) to develop an explainable deep learning algorithm for prediction of UIP from CT and validate its performance in independent cohorts. Methods: We trained a MIL algorithm using a pooled dataset (n=2,143) and tested it in three independent populations: data from a prior publication (n=127), a single-institution clinical cohort (n=239), and a national registry of patients with pulmonary fibrosis (n=979). We tested UIP classification performance using receiver operating characteristic (ROC) analysis with histologic UIP as ground truth. Cox proportional hazards and linear mixed effects models were used to examine associations between MIL predictions and survival or longitudinal forced vital capacity (FVC). Measurements and Main Results: In two cohorts with biopsy data, MIL improved accuracy for histologic UIP (area under the curve [AUC] 0.77 [n=127] and 0.79 [n=239]) compared to visual assessment (AUC 0.65 and 0.71). In cohorts with survival data, MIL UIP classifications were significant for mortality ([n=239, mortality to April 2021] unadjusted hazard ratio 3.1 95% confidence interval [CI] [1.96, 4.91] p<0.001, and [n=979, mortality to July 2022] 3.64 95% CI [2.66, 4.97] p<0.001). Individuals classified as UIP positive by the algorithm had a significantly greater annual decline in FVC than those classified as UIP negative (-88 ml/year versus -45 ml/year, n=979 p<0.01), adjusting for extent of lung fibrosis. Conclusions: Computerized assessment using MIL identifies clinically significant features of UIP on CT. Such a method could improve confidence in radiologic assessment of patients with interstitial lung disease, potentially enabling earlier and more precise diagnosis.