RESUMO
Mild cognitive impairment (MCI) is an unstable cognitive impairment state between normal aging and Alzheimer's disease (AD). The symptoms of MCI are mild and it has four different types of outcomes: reversing normal, maintaining stability, progression and death. However, 2/3 of MCI patients may still progress to dementia. Therefore, early identification of stable MCI (sMCI) and progressive MCI (pMCI) is beneficial for timely intervention, and delaying the progression of MCI, then improving patients' quality of life. Structural magnetic resonance imaging (sMRI) can predict dementia related neurodegeneration and cognitive decline. A large number of studies have found that, in addition to abnormalities in clinical scales, there are significant changes in sMRI during the progression of sMCI to pMCI, mainly including differences in cortical thickness and brain atrophy, hippocampal volume, and structural brain network connectivity. Especially, machine learning methods such as big data based neural convolutional networks are helpful in early prediction of sMCI and pMCI. These studies contribute to the discovery of early imaging markers for the conversion of sMCI to pMCI.