Purpose/Significance To Utilize deep neural networks to carry out the risk assessment study of lung cancer in the elderly, and to provide the theoretical basis for reducing the incidence of lung cancer and its effective preventive control. Method/Process A variety of related risk factors are integrated, and the integrated data are preprocessed. The deep neural network-based risk assessment model for lung cancer is constructed. And the comparative analysis is used to carry out the model validation evaluation. Result/Conclusion The incidence of lung cancer of the high-risk group is significantly higher than the low-risk group. The incidence of lung cancer of the male high-risk group is 2.02 times higher than the female high-risk group. The intervention control of the male high-risk group is more significant for reducing lung cancer incidence. The study is not only beneficial to the early diagnosis and treatment of lung cancer in the elderly, but also provides basic support for improving the quality of life of the elderly.