文献速递腹部技术与安全MR

基于T1和T2弛豫图的纹理分析联合机器学习算法以肝组织学为参考标准对肝纤维化分级的诊断准确性

Diagnostic accuracy of texture analysis applied to T1- and T2-Relaxation maps for liver fibrosis classification via machine-learning algorithms with liver histology as reference standard.
PubMed 文献 · Eur J Radiol · 2024-12-20

本研究回顾性纳入46例有肝组织学检查的3T MRI患者,提取T1与T2弛豫图的纹理特征,用支持向量机结合主成分分析分别鉴别无至轻度肝纤维化(nmF)与重度纤维化(sF)。TA-T1的AUC为0.748,与MRE的0.759相当(p=0.905);联合MRE与TA-T1的AUC为0.817,高于单用MRE的0.759但未达统计学显著(p=0.373)。作者建议将T1 mapping纳入肝脏临床与科研检查。

要点

  • T1弛豫图纹理分析(TA-T1)鉴别nmF与sF的AUC为0.748,与磁共振弹性成像(MRE)的0.759相当(p=0.905)。
  • MRE的AUC为0.759,优于T2的0.552和TA-T2的0.515;T1的AUC为0.692,略低于MRE但无统计学显著差异(p=0.396)。
  • 联合MRE与TA-T1的AUC为0.817,高于单用MRE的0.759但差异无统计学意义(p=0.373),提示两者可能互补。

核心结论

  • TA-T1在鉴别nmF与sF组方面与MRE表现相当,虽无统计学显著差异,但与MRE联合可提高分类性能,提示两者具有互补性;鉴于T1 mapping的广泛可用性、稳健性和短扫描时间,作者主张将其纳入每项临床和科研肝脏检查。

研究设计与样本

  • 单中心回顾性研究,纳入46例接受3T MRI且有组织学检查的患者,采用SNAPSHOT-FLASH序列行T1 mapping、径向涡轮自旋回波序列行T2 mapping、自旋回波平面成像磁共振弹性成像(MRE)序列,以组织学为参考标准。

诊断标准与关键数值

  • TA-T1鉴别nmF与sF的AUC为0.748,MRE的AUC为0.759(p=0.905)。
  • T1鉴别nmF与sF的AUC为0.692,与MRE的0.759相比无统计学显著差异(p=0.396)。
  • MRE的AUC为0.759,高于T2的AUC 0.552和TA-T2的AUC 0.515。
  • RFECV算法识别出四个特征:MRE、T1及前两个TA-T1主成分,构成第一联合模型。
  • 第一联合模型(MRE+T1+TA-T1)AUC为0.797,与MRE的0.759相比无统计学显著差异(p=0.597);第二联合模型(MRE+TA-T1)AUC为0.817,与MRE的0.759相比无统计学显著差异(p=0.373)。

临床应用与阅片要点

  • 作者主张将T1 mapping纳入每项临床和科研肝脏检查。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

英文原文(摘要)· 点击展开

OBJECTIVES: To explore texture analysis' ability on T1 and T2 relaxation maps to classify liver fibrosis into no-to-mild liver fibrosis (nmF) versus severe fibrosis (sF) group using machine learning algorithms and histology as reference standard.

MATERIALS AND METHODS: In this single-center study, patients undergoing 3 T MRI who also had histology examination were retrospectively enrolled. SNAPSHOT-FLASH sequence for T1 mapping, radial turbo-spin-echo sequence for T2 mapping and spin-echo echo-planar-imaging magnetic resonance elastography (MRE) sequences were analyzed. Grey-level co-occurrence matrix texture analysis features were extracted from T1 (TA-T1) and T2 (TA-T2) maps from single-slice whole-liver region-of-interest. The extracted features were evaluated as predictors for nmF and sF group classification separately using support-vector-machine algorithm combined with principal component analysis in case of texture features. Recursive Feature Elimination with Cross-Validation (RFECV) was used to identify the most significant features and the importance of selected features was assessed with permutation importance algorithm. A combined model was identified and evaluated. Area under the receiver operating characteristic curve (AUC) was used for scoring and model comparison.

RESULTS: 46 patients (mean age 52.8 ± 16.1 years, 23 males) were evaluated. TA-T1 performed comparably to MRE (0.748 vs 0.759, p = 0.905) and T1 performed slightly worse compared to MRE which was not statistically significant (0.692 vs 0.759, p = 0.396). MRE outperformed T2 (0.759 vs 0.552) and TA-T2 (0.759 vs 0.515). RFECV algorithm identified four features: MRE, T1 and 1st two TA-T1 principal components, constituting the first combined model. The permutation importance identified T1 as feature of very low importance, therefore a second combined model was constructed, omitting T1 from the first combined model. Even though both combined models performed higher than MRE (0.759 vs 0.797, p = 0.597 for MRE vs MRE + T1 + TA-T1, and 0.759 vs 0.817, p = 0.373 for MRE vs MRE + TA-T1), it was not statistically significant.

CONCLUSIONS: TA-T1 performed comparably to MRE in liver fibrosis classification to nmF and sF groups, and even though not statistically significant, combining those with MRE increased the performance, suggesting their complementary nature. Given the broad availability, robustness and short scanning times of T1 mapping, we would advocate for the inclusion of T1 mapping in every clinical and research liver examination.

来源:Eur J Radiol,查看原文。版权归原作者所有。

引用

  1. Sitarcikova D, Poetter-Lang S, Bastati N, et al. Diagnostic accuracy of texture analysis applied to T1- and T2-Relaxation maps for liver fibrosis classification via machine-learning algorithms with liver histology as reference standard. Eur J Radiol. 2024. PMID: 39721334. doi:10.1016/j.ejrad.2024.111887

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。