Basukala Dibash, Mikheev Artem, Li Xiaochun, Goldberg Judith D, Gilani Nima, Moy Linda, Pinker Katja, Partridge Savannah C, Biswas Debosmita, Kataoka Masako, Honda Maya, Iima Mami, Thakur Sunitha B, Sigmund Eric E
Department of Radiology, Grossman School of Medicine, New York University, New York, NY, United States.
Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, United States.
Front Oncol. 2025 Feb 24;15:1524634. doi: 10.3389/fonc.2025.1524634. eCollection 2025.
The intravoxel incoherent motion (IVIM) model of diffusion weighted imaging (DWI) provides imaging biomarkers for breast tumor characterization. It has been extensively applied for both diagnostic and prognostic goals in breast cancer, with increasing evidence supporting its clinical relevance. However, variable performance exists in literature owing to the heterogeneity in datasets and quantification methods.
This work used retrospective anonymized breast MRI data (302 patients) from three sites employing three different software utilizing least-squares segmented algorithms and Bayesian fit to estimate 1 order radiomics of IVIM parameters perfusion fraction ( ), pseudo-diffusion ( ) and tissue diffusivity ( ). Pearson correlation () coefficients between software pairs were computed while logistic regression model was implemented to test malignancy detection and assess robustness of the IVIM metrics.
and maps generated from different software showed consistency across platforms while maps were variable. The average correlation between the three software pairs at three different sites for 1 order radiomics of IVIM parameters were /////: 0.791/0.891/0.98/0.815/0.697/0.584; ////: 0.615/0.871/0.679/0.541/0.433; ////: 0.616/0.56/0.587/0.454/0.51. Correlation between least-squares algorithms were the highest. showed highest area under the ROC curve (AUC) with 0.85 and lowest coefficient of variation (CV) with 0.18% for benign and malignant differentiation using logistic regression. metrics were highly diagnostic as well as consistent along with metrics.
Multiple 1 order radiomic features of and obtained from a heterogeneous multi-site breast lesion dataset showed strong software robustness and/or diagnostic utility, supporting their potential consideration in controlled prospective clinical trials.
扩散加权成像（DWI）的体素内不相干运动（IVIM）模型可为乳腺肿瘤特征提供成像生物标志物。它已广泛应用于乳腺癌的诊断和预后评估，越来越多的证据支持其临床相关性。然而，由于数据集和量化方法的异质性，文献中的表现存在差异。
本研究使用了来自三个地点的302例患者的回顾性匿名乳腺MRI数据，采用三种不同软件，利用最小二乘分割算法和贝叶斯拟合来估计IVIM参数灌注分数（）、伪扩散（）和组织扩散率（）的一阶影像组学特征。计算软件对之间的Pearson相关（）系数，同时实施逻辑回归模型以测试恶性肿瘤检测并评估IVIM指标的稳健性。
不同软件生成的和图在各平台间显示出一致性，而图则存在差异。三个不同地点的三种软件对IVIM参数一阶影像组学特征的平均相关性为：/////: 0.791/0.891/0.98/0.815/0.697/0.584；////: 0.615/0.871/0.679/0.541/0.433；////: 0.616/0.56/0.587/0.454/0.51。最小二乘算法之间的相关性最高。使用逻辑回归进行良恶性鉴别时，显示出最高的ROC曲线下面积（AUC）为0.85，最低的变异系数（CV）为0.18%。指标具有高度诊断性，并且与指标一致。
从异质性多中心乳腺病变数据集中获得的多个和的一阶影像组学特征显示出强大的软件稳健性和/或诊断效用，支持在对照前瞻性临床试验中对其进行潜在考虑。
