The Alzheimer's disease (AD) is usually insidious in its onset and there are no drugs or methods to effectively control and treat the disease. Early prediction and intervention during the stage of the mild cognitive impairment (MCI) can effectively delay the course of the disease. The review contains two aspects. One is the early clinical diagnosis of the AD based on the brain imaging radiomics features, another is the AD early prediction based on the artificial intelligence (AI) of the imaging radiomics. We propose that in the framework of the deep learning, the imaging and the genomics are combined to construct a deep learning model with a high classification and prediction performance, to provide support for early screening and intervention of the AD.

radiomics  /  artificial intelligence  /  Alzheimer's disease