人工智能在脑胶质瘤分子标志物无创诊断研究中的应用进展
Advances in artificial intelligence in the non-invasive diagnosis of molecular markers of glioma
胡平 1江洪祥 2王龙 2邓钢2
作者信息
- 1. 武汉大学人民医院神经外科,武汉 430060;南昌大学第二附属医院神经外科,南昌 330006
- 2. 武汉大学人民医院神经外科,武汉 430060
- 折叠
摘要
早期诊断脑胶质瘤分子标志物对其治疗和预后至关重要,基于人工智能算法在脑胶质瘤MR图像中提取的高通量影像特征已成为预测胶质瘤分子病理亚型的潜在无创生物标志物,开启了胶质瘤"分子成像"的时代.本研究旨在对近年来应用人工智能算法在脑胶质瘤MR图像中的病灶分割和构建影像组学或影像基因组学等模型非侵入性诊断分子标志物的文献进行综述,以期确定人工智能能够对胶质瘤分子生物标志物进行无创预测,对辅助临床医生作出决策提供可靠的依据.
Abstract
Early diagnosis of molecular markers of glioma is crucial for its treatment and prognosis,and high-throughput imaging features extracted from MR images of glioma based on artificial intelligence algorithms have become potential noninvasive biomarkers for predicting the molecular pathological subtype of glioma,opening the era of"molec-ular imaging"of glioma.This study aims to review the literature on the application of artificial intelligence algorithms in lesion segmentation in MR images of glioma and the construction of non-invasive diagnostic molecular markers such as radiomics or imaging genomics,in order to determine that artificial intelligence can make non-invasive predictions of glioma molecular biomarkers and provide a reliable basis for assisting clinicians in making decisions.
关键词
人工智能/影像组学/影像基因组学/脑胶质瘤/分子病理Key words
Artificial intelligence/Radiomics/Radiogenomics/Glioma/Molecular pathology引用本文复制引用
基金项目
武汉市知识创新专项曙光计划(2022020801020483)
武汉大学本科教育质量建设综合改革项目(2022ZG188)
出版年
2024
NETL
NSTL
万方数据