阿尔茨海默病影像组学关键方法研究进展

Research progresses in key methods of radiomics for Alzheimer disease

随着医学成像技术的发展,基于影像组学对阿尔茨海默病(AD)进行研究已成为当前热点之一。本文对现有影像组学方法在AD中的应用研究进展做一综述。首先阐述基于机器学习的影像组学和基于深度学习的影像组学方法基本工作流程;其次对基于机器学习的影像组学特征提取、选择和降维方法及统计分类模型进行概述;而后讨论近期基于深度学习的影像组学方法在AD中的应用;最后对比分析机器学习与深度学习方法在实际应用中存在的局限性与挑战。

With the development of medical imaging technology,research of Alzheimer disease(AD)based on radiomics has become one of the current hot spots.The applications of existing radiomics analysis methods in AD were reviewed in this article.Firstly,the workflows of radiomics with machine learning and deep learning were described.Secondly,the key methods about feature extraction,selection,dimensionality reduction and statistical classification models of radiomics with machine learning were summarized.Then the applications of recent deep learning-based radiomics method in AD were discussed.Finally,the limitations and challenges of machine learning and deep learning methods in practical applications were compared and analyzed.