The paper introduces the influence of confounding factors on causal inference in medical studies and the identification methods of common confounding factors, summarizes the application of machine learning methods for controlling confounding factors in causal inference, and discusses the opportunities and challenges of applying machine learning methods to control confounding factors.