基于T2WI影像组学鉴别肝脏局灶性病变的研究

A Radiomics Model Based on T2WI Sequence to Identify Focal Liver Lesions

目的:旨在探讨T2WI序列的影像组学模型鉴别HCC和FNH的价值。方法:回顾性分析2011年至2021年间在福建医科大学附属漳州市医院确诊的196名患者的T2WI序列图像,对T2WI序列的影像组学特征提取后,通过方差阈值法、K最佳方法和Lasso回归操作进行特征筛选和降维,采用KNN分类器构建了HCC和FNH鉴别诊断模型。结果:该模型在测试组的结果是AUC值为0.823,准确率为0.750,敏感度0.933,特异性0.567。结论:本研究采用T2WI序列结合影像组学方法构建了HCC和FNH的鉴别诊断模型,该模型显示出较高的敏感性及AUC值。这一结果突显了利用单一T2WI序列提取影像组学特征在鉴别肝脏病变方面的潜力,展示了影像组学在提高非侵入性诊断精度方面的应用前景,同时也为探索多种MR序列及临床数据相互融合的综合模型提供研究基础。

Objective:To investigate the value of radiomics model based on T2WI sequence in differentiating HCC from FNH.Methods:A retrospective analysis was conducted on the T2WI sequence images of 196 patients who were diagnosed at Zhangzhou Affiliated Hospital of Fujian Medical University between 2011 and 2021.After extracting the radiomics features of T2WI sequence,the variance threshold method,K best method and Lasso regression operation were used for feature selection and dimensionality reduction.The KNN classifier was used to construct a differential diagnosis model for HCC and FNH.Results:The AUC of the model in the test group was 0.823,the accuracy was 0.750,the sensitivity was 0.933,and the specificity was 0.567.Conclusion:In this study,a model based on T2WI sequence combined with radiomics was established for the differential diagnosis of HCC and FNH,which showed high sensitivity and AUC.This finding highlights the potential of radiomics features extracted from a single T2WI sequence to identify liver lesions,demonstrates the application prospect of radiomics in improving the accuracy of non-invasive diagnosis,and provides a research basis for exploring a comprehensive model based on the integration of multiple MR Sequences and clinical data.