Bone tumors have a wide variety and complex classification,even some of them have similar imaging manifesta-tions.For example,a few benign bone tumors may show invasive characteristics similar to the malignant bone tumors,which makes the diagnosis and differential diagnosis of bone tumors based on traditional imaging facing great challenges.In recent years,the rapid development of radiomics and artificial intelligence(AI)technology has greatly improved the ability of lesion image recognition,which is widely used in the classification and characterization of various tumors.This article mainly reviews the advances in radiomics of bone tumor diagnosis,prognosis prediction and genetic analysis and prediction,and puts forward the current challenges and future development direction.
bone tumorradiomicsfeature extractionmachine learning
NETL
NSTL
万方数据
