影像组学在胃肠道间质瘤诊疗中的研究进展

Research progress of radiomics in the diagnosis and treatment of gastrointestinal stromal tumors

胃肠道间质瘤(GIST)是胃肠道最主要的间叶源性肿瘤,其生物学特征复杂,风险程度各异,因此早期诊断及精准的危险度评估对于后续的治疗至关重要。影像组学可以从不同模态的影像数据中提取并分析具有强代表性的定量影像特征,通过机器学习的方法结合临床数据,完成对病变诊断和预测等工作。现有研究表明,影像组学不仅能用于GIST的鉴别诊断、危险度分层及预后判断,还在基因突变状态评估和治疗反应预测方面展现出潜力。本文就影像组学在GIST早期鉴别诊断、危险度分级、基因突变预测、治疗效果评估等方面展开综述,并报告其局限性,探讨未来的发展方向。

Gastrointestinal stromal tumor(GIST)is the most common mesenchymal tumor of the gastrointestinal tract,with complex biological characteristics and varying risk levels,thus early diagnosis and precise risk assessment are crucial for subsequent treatment.Imaging genomics can extract and analyze quantitative imaging features with strong representativeness from different modalities of imaging data,and complete the tasks of diagnosis and prediction by combining machine learning methods with clinical data.Studies have shown that imaging genomics can not only be used for differential diagnosis,risk stratification,and prognosis assessment of GIST,but also show potential in gene mutation status assessment and treatment response prediction.This review discussed the application of imaging genomics in early differential diagnosis,risk grading,gene mutation prediction,and treatment effect evaluation of GIST,and reported its limitations,in order to explore the future development directions.