首页|灌注加权成像预测成人胶质瘤预后的生物学基础

灌注加权成像预测成人胶质瘤预后的生物学基础

扫码查看
目的 探讨动态敏感性对比灌注加权成像(DSC-PWI)在预测成人弥漫性胶质瘤预后中的价值及背后的生物学意义。方法 选取2013-06-2021-07 郑州大学第一附属医院神经外科收治并经病理检查结果确诊的343 例成人型弥漫性胶质瘤患者。收集患者术前PWI影像数据及肿瘤标本,随机分为训练集(175 例)和测试集(168 例)。从PWI影像中提取影像组学特征,经筛选后建立定性的影像学预测模型。基于影像学特征及RNA-seq基因组学分析,通过基因探针富集分析(GSEA)分析和加权基因共表达网络分析(WGCNA)统合识别生物学通路。结果 筛选出7 个影像组学特征,计算出相应的评分,证明是预测胶质瘤预后的因素(P<0。05)。同时发现了缺氧诱导、免疫调节及肿瘤增殖三种与影像临床复合模型明显相关的生物学通路类别。结论 本研究将基于机器学习的方法构建了一种新型联合模型,对胶质瘤展示出较好的预测性能。再结合RNA-seq数据找到了与胶质瘤有关的三种生物学通路。
Biological basis of perfusion weighted imaging for predicting prognosis of adult gliomas
Objective To explore the value and biological significance of dynamic sensitivity contrastive perfusion weighted imaging(DSC-PWI)in predicting the prognosis of diffuse glioma in adults.Methods From June 2013 to July 2021,totally 343 patients with adult diffuse gliomas admitted to Neurosurgery of the First Hospital Affiliated to Zhengzhou University and confirmed by pathological examination were selected.Collect preoperative PWI imaging data and tumor samples from patients,and randomly divided them into a training set(n=175)and a testing set(n=168).Extract imaging omics features from PWI images,and establish a qualitative imaging prediction model after screening.Based on imaging characteristics and RNA seq Genomics analysis,biological pathways were identified through gene probe enrichment analysis(GSEA)and weighted Gene co-expression network analysis(WGCNA).Results Seven imaging omics features were selected and corre-sponding scores were calculated,proving to be a predictive factor for the prognosis of glioma(P<0.05).At the same time,three biological path-way categories were found to be significantly related to the imaging clinical composite model,including hypoxia induction,immune regulation,and tumor proliferation.Conclusion This study will construct a new joint model based on machine learning methods,demonstrating good predictive performance for gliomas.Combined with RNA seq data,three biological pathways related to glioma were identified.

GliomaMachine learningImage GenomicsPWIRNA seqGSEAWGCNA

郭语、郑惠敏、刘献志

展开 >

郑州大学第一附属医院神经外科 郑州 450052

胶质瘤 机器学习 影像基因组学 PWI RNA-seq GSEA WGCNA

2024

河南外科学杂志
郑州大学

河南外科学杂志

影响因子:0.709
ISSN:1007-8991
年,卷(期):2024.30(3)
  • 33