机器学习在神经精神疾病诊断及预测中的应用

Application of Machine Learning in Diagnosis and Prediction of Neuropsychiatric Diseases

多模态磁共振神经影像技术的发展为大脑工作原理研究及脑部疾病早期诊断提供了新的手段。当前多数神经精神性疾病的诊断仅依据其临床症状,缺少客观的神经影像生物学标志物。传统的基于组间比较的单变量分析仅能在组间水平进行推断,无法提供个体水平的诊断和预测,临床应用价值非常有限。机器学习技术可在不同水平对神经影像进行计算分析和研究,发现其中规律从而有效预测和分类未知数据,找出与脑疾病高度相关的脑区特征,提供个体水平的诊断并探测脑疾病的病理生理机制。本文对基于机器学习的神经影像分析步骤及机器学习在神经精神疾病智能诊断及预测研究中的应用进行综述,并对未来研究发展进行展望。

The development of multi-modal magnetic resonance imaging （MRI） provides a new method for early diagnosis of brain diseases. Currently, diagnosis of most neuropsychiatric diseases is only based on clinical symptoms,which lacks objective neuroimaging biomarkers. As univariate analysis that is used in traditional research can only reveal disease-related structural and functional alterations at group level, which limits the clinical application. Recent attention has turned toward integrating multi-modal neuroimaging and machine learning to assist clinical disease diagnosis. Machine learning can obtain rules via automatically analyzing neuroimaging data, and apply these rules to predict unknown data, to identify the brain regions highly correlated to the brain disease, to provide individual levels of diagnosis, and to detect pathophysiological mechanisms. This paper reviews the concrete steps of neuroimaging analysis based on machine learning and the application of machine learning in intelligent diagnosis and prediction of neuropsychiatric diseases. Finally, some new directions for future research are forecasted.