Development and validation of a clinically applicable deep learning strategy (HONORS) for pulmonary nodule classification at CT: A retrospective multicentre study.

PURPOSE: To propose a practical strategy for the clinical application of deep learning algorithm, i.e., Hierarchical-Ordered Network-ORiented Strategy (HONORS), and a new approach to pulmonary nodule classification in various clinical scenarios, i.e., Filter-Guided Pyramid NETwork (FGP-NET). MATERIALS AND METHODS: We developed and validated FGP-NET on a collection of 2106 pulmonary nodules on computed tomography images which combined screened and clinically detected nodules, and performed external test (n = 341). The area under the curves (AUCs) of FGP-NET were assessed. A comparison study with a group of 126 skilled radiologists was conducted. On top of FGP-NET, we built up our HONORS which was composed of two solutions. In the Human Free Solution, we used the high sensitivity operating point for screened nodules, but the high specificity operating point for clinically detected nodules. In the Human-Machine Coupling Solution, we used the Youden point. RESULTS: FGP-NET achieved AUCs of 0.969 and 0.847 for internal and external test. The AUCs of the subsets of the external test set ranged from 0.890 to 0.942. The average sensitivity and specificity of the 126 radiologists were 72.2 ±â€¯15.1 % and 71.7 ±â€¯15.5 %, respectively, while a higher sensitivity (93.3 %) but a relatively inferior specificity (64.0 %) were achieved by FGP-NET. HONORS-guided FGP-NET identified benign nodules with high sensitivity (sensitivity,95.5 %; specificity, 72.5 %) in the screened nodules, and identified malignant nodules with high specificity (sensitivity, 31.0 %; specificity, 97.5 %) in the clinically detected nodules. These nodules could be reliably diagnosed without any intervention from radiologists, via the Human Free Solution. The remaining ambiguous nodules were diagnosed with high performance, which however required manual confirmation by radiologists, via the Human-Machine Coupling Solution. CONCLUSIONS: FGP-NET performed comparably to skilled radiologists in terms of diagnosing pulmonary nodules. HONORS, due to its high performance, might reliably contribute a second opinion, aiding in optimizing the clinical workflow.

Discrimination between transient and persistent subsolid pulmonary nodules on baseline CT using deep transfer learning.

OBJECTIVES: To develop and validate a deep learning model to discriminate transient from persistent subsolid nodules (SSNs) on baseline CT. METHODS: A cohort of 1414 SSNs, consisting of 319 transient SSNs in 168 individuals and 1095 persistent SSNs in 816 individuals, were identified on chest CT. The cohort was assigned by examination date into a development set of 996 SSNs, a tuning set of 212 SSNs, and a validation set of 206 SSNs. Our model was built by transfer learning, which was transferred from a well-performed deep learning model for pulmonary nodule classification. The performance of the model was compared with that of two experienced radiologists. Each nodule was categorized by Lung CT Screening Reporting and Data System (Lung-RADS) to further evaluate the performance and the potential clinical benefit of the model. Two methods were employed to visualize the learned features. RESULTS: Our model achieved an AUC of 0.926 on the validation set with an accuracy of 0.859, a sensitivity of 0.863, and a specificity of 0.858, and outperformed the radiologists. The model performed the best among Lung-RADS 2 nodules and maintained well performance among Lung-RADS 4 nodules. Feature visualization demonstrated the model's effectiveness in extracting features from images. CONCLUSIONS: The transfer learning model presented good performance on the discrimination between transient and persistent SSNs. A reliable diagnosis on nodule persistence can be achieved at baseline CT; thus, an early diagnosis as well as better patient care is available. KEY POINTS: • Deep learning can be used for the discrimination between transient and persistent subsolid nodules. • A transfer learning model can achieve good performance when it is transferred from a model with a similar task. • With the assistance of deep learning model, a reliable diagnosis on nodule persistence can be achieved at baseline CT, which can bring a better patient care strategy.