影像组学与肿瘤风险评估、预后和疗效预测的研究进展

Research progress in radiomics for tumor risk assessment, prognosis, and therapeutic response prediction

关键词:  影像组学  机器学习  肿瘤研究

Key words:  radiomics  machine learning  tumor research

摘要:

影像组学通过提取分析影像图片的定量特征建立机器学习模型在肿瘤诊断、治疗和监测过程中表现出巨大潜力。近来影像组学研究已经初具规模,在方法流程上也在不断成熟和完善,但距离临床应用还存在一定困难。笔者从实用性角度简要概况了影像组学的各个流程及相应的分析工具和方法,并以案例的形式重点介绍影像组学在肿瘤危险分层、预后预测和疗效预测等研究中的应用、不足和难点,以期为肿瘤的精准诊疗提供新的技术和方法。

Abstract:

Radiomics, by extracting and analyzing quantitative features of images and constructing machine learning models, has shown great potential in the process of tumor diagnosis, treatment and monitoring. Recently, radiomics research has begun to take shape, and the methods and procedures are also maturing and improving, but there are still some difficulties in clinical application. From the perspective of practicality, the author briefly summarizes the various processes of radiomics and the corresponding analytical tools and methods, and focuses on the application, shortcomings and difficulties of radiomics in the study of tumor risk stratification, prognosis prediction and curative effect prediction in the form of cases, with a view to providing new technologies and methods for accurate diagnosis and treatment of tumors.