In the past decades, the quantity and complexity of medical data have been increasing, such as medical images, genomics data, physiological signals and so on. These data contain a lot of valuable information, but traditional analysis methods and manual feature extraction have been unable to effectively deal with these large-scale and high-dimensional data. The rapid development of machine learning technology has brought new opportunities for medical research and clinical practice. Machine learning algorithms can learn and discover patterns, laws and prediction models from large-scale data, thus helping doctors and researchers to make more accurate and personalized diagnosis and treatment decisions. The application of machine learning in medicine has become a research field of great concern in recent years. This paper studies the application of machine learning in medicine, such as medical image diagnosis, genomics and drug discovery, and analyzes the relevant technical methods and ideas of machine learning in medicine, and analyzes the main algorithms and usage methods used in medical image diagnosis, such as convolution neural network and other deep learning algorithms. Through detailed analysis and research, it is found that machine learning provides a new method and tool, which can effectively process large-scale and complex medical data in medicine, bring more possibilities for medical diagnosis, treatment and research, and provide support for individualized medical care.