Radiomics-based decision support tool assists radiologists in small lung nodule classification and improves lung cancer early diagnosis

Abstract

Background: Methods to improve stratification of small (≤15 mm) lung nodules are needed. We aimed to develop a radiomics model to assist lung cancer diagnosis.

Methods: Patients were retrospectively identified using health records from January 2007 to December 2018. The external test set was obtained from the national LIBRA study and a prospective Lung Cancer Screening programme. Radiomics features were extracted from multi-region CT segmentations using TexLab2.0. LASSO regression generated the 5-feature small nodule radiomics-predictive-vector (SN-RPV). K-means clustering was used to split patients into risk groups according to SN-RPV. Model performance was compared to 6 thoracic radiologists. SN-RPV and radiologist risk groups were combined to generate "Safety-Net" and "Early Diagnosis" decision-support tools.

Results: In total, 810 patients with 990 nodules were included. The AUC for malignancy prediction was 0.85 (95% CI: 0.82-0.87), 0.78 (95% CI: 0.70-0.85) and 0.78 (95% CI: 0.59-0.92) for the training, test and external test datasets, respectively. The test set accuracy was 73% (95% CI: 65-81%) and resulted in 66.67% improvements in potentially missed [8/12] or delayed [6/9] cancers, compared to the radiologist with performance closest to the mean of six readers.

Conclusions: SN-RPV may provide net-benefit in terms of earlier cancer diagnosis.

Fig. 1. Recruitment diagram and segmentation labels.

a Study recruitment diagram. Databases of lung histopathology (n = 1450) and lung nodule MDT records (n = 927) were used to identify eligible patients. Following exclusion based on eligibility criteria (n = 1723) and technical limitations of CT images (n = 21), the final internal dataset consisted of 633 patients with 736 nodules. b Cropped, axial plane CT images showing binary segmentation masks (red) for nodule regions. The primary lung nodule was segmented (1) and then expanded by 2 mm isotropically to create a spherical annulus structure (2). An 8 × 8 mm spherical background structure (3) was segmented 15 mm away from the primary lesion.

Fig. 2. Decision-support tool scenarios.

In cases where the radiologist categorises a nodule as low-risk, a high-risk SN-RPV triggers the “Safety Net”, prompting earlier surveillance or investigation. In cases where the radiologist classifies a nodule as indeterminate, a high-risk SN-RPV triggers the “Early Diagnosis” pathway, prompting further investigation rather than surveillance. Cases evaluated as high-risk by the radiologist are not affected by the proposed decision support method.