Application of Neural Network Algorithm in Medical Artificial Intelligence Product Development

With the continuous deepening of artificial intelligence (AI) in the medical field, the social risks brought by the development and application of medical AI products have become increasingly prominent, bringing hidden worries to the protection of civil rights, social stability, and healthy development. There are many new problems that need to be solved in our country's existing risk regulation theories when dealing with such risks. By introducing the theory of risk administrative law, it analyzes the social risks of medical AI, organically combines the principle of risk prevention with benefit measurement, and systematically and flexibly reconstructs the theoretical system of medical AI social risk assessment. This paper has completed the following work: (1) reviewed and sorted out the works and papers related to medical AI ethics, medical AI risk, etc., and sorted out the current situation of medical AI social risk regulation at home and abroad to provide help for follow-up research. (2) The related technologies of artificial neural network (ANN) are introduced, and the risk assessment index system of medical AI is constructed. (3) With the self-designed dataset, the trained neural network model is utilized to assess risk. The experimental results reveal that the created BPNN model's error is relatively tiny, indicating that the algorithm model developed in this research is worth popularizing and applying.