Development and validation of a deep learning model to predict survival of patients with esophageal cancer

OBJECTIVE: To compare the performance of a deep learning survival network with the tumor, node, and metastasis (TNM) staging system in survival prediction and test the reliability of individual treatment recommendations provided by the network. METHODS: In this population-based cohort study, we developed and validated a deep learning survival model using consecutive cases of newly diagnosed stage I to IV esophageal cancer between January 2004 and December 2015 in a Surveillance, Epidemiology, and End Results (SEER) database. The model was externally validated in an independent cohort from Fujian Provincial Hospital. The C statistic was used to compare the performance of the deep learning survival model and TNM staging system. Two other deep learning risk prediction models were trained for treatment recommendations. A Kaplan–Meier survival curve was used to compare survival between the population that followed the recommended therapy and those who did not. RESULTS: A total of 9069 patients were included in this study. The deep learning network showed more promising results in predicting esophageal cancer-specific survival than the TNM stage in the internal test dataset (C-index=0.753 vs. 0.638) and external validation dataset (C-index=0.687 vs. 0.643). The population who received the recommended treatments had superior survival compared to those who did not, based on the internal test dataset (hazard ratio, 0.753; 95% CI, 0.556-0.987; P=0.042) and the external validation dataset (hazard ratio, 0.633; 95% CI, 0.459-0.834; P=0.0003). CONCLUSION: Deep learning neural networks have potential advantages over traditional linear models in prognostic assessment and treatment recommendations. This novel analytical approach may provide reliable information on individual survival and treatment recommendations for patients with esophageal cancer.