Rihui Zhang ¹ Haiming Zhu ^1* Minbin Chen ² Weiwei Sang ^1* Ke Lu ^3* Zhen Li ⁴ Chunhao Wang ³ Lei Zhang ¹ Fangfang Yin ¹ Zhenyu Yang ^1*

• ¹ Duke Kunshan University, Kunshan, China
• ² First People's Hospital of Kunshan, Kunshan, Jiangsu, China
• ³ Department of Radiation Oncology, School of Medicine, Duke University, Durham, North Carolina, United States
• ⁴ Shanghai Sixth People's Hospital, Shanghai Jiao Tong University, Shanghai, Shanghai Municipality, China

Radiation therapy (RT) is one of the primary treatment options for early-stage non-small cell lung cancer (ES-NSCLC). Therefore, accurately predicting the overall survival (OS) rate following radiotherapy is crucial for implementing personalized treatment strategies. This work aims to develop a dual-radiomics (DR) model to (1) predict 3-year OS in ES-NSCLC patients receiving RT using pretreatment CT images, and (2) provide explanations between feature importance and model prediction performance. The publicly available TCIA Lung1 dataset with 132 ES-NSCLC patients received RT were studied: 89/43 patients in the under/over 3-year OS group. For each patient, two types of radiomic features were examined: 56 handcrafted radiomic features (HRFs) extracted within gross tumor volume, and 512 image deep features (IDFs) extracted using a pre-trained U-Net encoder. They were combined as inputs to an explainable boosting machine (EBM) model for OS prediction. The EBM's mean absolute scores for HRFs and IDFs were used as feature importance explanations. To evaluate identified feature importance, the DR model was compared with EBM using either (1) key or (2) non-key feature type only. Comparison studies with other models, including supporting vector machine (SVM) and random forest (RF), were also included. The performance was evaluated by the area under the receiver operating characteristic curve (AUCROC), accuracy, sensitivity, and specificity with a 100-fold Monte Carlo cross-validation. The DR model showed highest performance in predicting 3-year OS (AUCROC=0.81±0.04), and EBM scores suggested that IDFs showed significantly greater importance (normalized mean score=0.0019) than HRFs (score=0.0008). The comparison studies showed that EBM with key feature type (IDFs-only) demonstrated comparable AUCROC results (0.81±0.04), while EBM with non-key feature type (HRFs-only) showed limited AUCROC (0.64±0.10). The results suggested that feature importance score identified by EBM is highly correlated with OS prediction performance. Both SVM and RF models were unable to explain key feature type while showing limited overall AUCROC=0.66±0.07 and 0.77±0.06, respectively. Accuracy, sensitivity, and specificity showed a similar trend. In conclusion, a DR model was successfully developed to predict ES-NSCLC OS based on pre-treatment CT images. The results suggested that the feature importance from DR model is highly correlated to the model prediction power.