基于多模态影像的机器学习与深度学习在帕金森病诊治中的应用进展

Progress on the application of machine learning and deep learning based on multimodal imaging in diagnosis and treatment of Parkinson's disease

帕金森病是一种常见的神经系统退行性疾病,在出现临床症状时已经不可逆,所以早期诊断显得尤为重要。随着人工智能技术的发展,机器学习与深度学习已广泛用于医学数据分析和神经影像自动化诊断研究。本文将综述基于多模态影像的机器学习与深度学习在帕金森病早期诊断、分类研究以及功能预测中的应用情况,并对其局限性作出阐述,以及对未来的发展趋势进行展望。

Parkinson's disease(PD)is a common neurodegenerative disease,it is irreversible when clinical symptoms appear,so early diagnosis is particularly important.With the development of artificial intelligence technology,machine learning and deep learning has been widely used in medical data analysis and automated neuroimaging diagnosis.This article reviews the application of machine learning and deep learning based on multimodal imaging in the early diagnosis,disease classification and functional prediction of Parkinson's disease,its limitations and future development trend are also discussed.