Currently, with the approval and clinical use of disease modifying therapy drugs such as lencanazumab and donenumab, mild cognitive impairment (MCI) and mild dementia stages have become the golden window for intervention in Alzheimer's disease patients. On the 7th, it was reported that Chinese medical experts have obtained the latest research results: they have successfully constructed a MCI prediction model, providing a new method for early recognition of cognitive impairment. It is reported that the research results obtained by the team of Professor Xiao Shifu/Yue Ling from the Mental Health Center affiliated with Shanghai Jiao Tong University School of Medicine, together with the team of Professor Shen Dinggang/Professor Pan Yongsheng from ShanghaiTech University, have been published in the well-known journal "Journal of Prevention of Alzheimer's Disease". It is reported that mild cognitive impairment (MCI) is considered a pre risk state for cognitive disorders such as Alzheimer's disease. This study is particularly crucial for early prediction of cognitive impairment: it not only helps identify potential patients, but also provides the possibility for timely and effective treatment, thereby delaying disease progression and improving patient prognosis. It is reported that the model is based on structural magnetic resonance imaging (MRI) data and has established a deep learning training framework; By using a network based on multiple regions of interest (MRNet) to screen and integrate 10 highly discriminative brain regions, including the hippocampus, amygdala, cerebellum, etc., a progression index (PI) was further constructed. This index is based on a continuous quantitative evaluation system constructed using deep learning. By analyzing brain MRI, a risk score with an interval of [0,1] is generated. The higher the score, the higher the risk of developing cognitive impairment in the future. PI breaks through the limitations of traditional binary diagnosis and can dynamically reflect the intermediate progression status of neurodegenerative diseases such as Alzheimer's disease. To further evaluate the predictive ability of PI for cognitive impairment risk, the research team combined demographic information, neuropsychological scales, and PI scores from MRI scans to establish an MCI prediction model that can effectively predict the transition to MCI in relevant cohorts after 7 years. It is reported that this study has received funding from the Major Project of Science and Technology Innovation 2030- "Neuroscience and Brain like Research" (2022ZD0213100). In the future, the research team will explore the potential of more cutting-edge deep learning paradigms in the clinical diagnosis and treatment of cognitive impairment related to Alzheimer's disease, providing more support for precise clinical diagnosis and treatment of mental illnesses. (New Society)