From the Institute for Experimental Molecular Imaging (Z.A.M., R.R., M.P., V.S., F.K.), Institute of Pathology (P.B.), and Department of Obstetrics and Gynecology (M.K., B.S.W., T.P., K.K., E.S.), University Clinic Aachen, RWTH Aachen University, Forckenbeckstrasse 55, 52074 Aachen, Germany; Physics Institute III B, RWTH Aachen University, Aachen, Germany (V.S.); Comprehensive Diagnostic Center Aachen, Uniklinik RWTH Aachen, Aachen, Germany (P.B., V.S., E.S., F.K.); and Fraunhofer Institute for Digital Medicine MEVIS, Bremen, Germany (P.B., V.S., F.K.).
Radiology. 2024 Sep;312(3):e232554. doi: 10.1148/radiol.232554.
Background US is clinically established for breast imaging, but its diagnostic performance depends on operator experience. Computer-assisted (real-time) image analysis may help in overcoming this limitation. Purpose To develop precise real-time-capable US-based breast tumor categorization by combining classic radiomics and autoencoder-based features from automatically localized lesions. Materials and Methods A total of 1619 B-mode US images of breast tumors were retrospectively analyzed between April 2018 and January 2024. nnU-Net was trained for lesion segmentation. Features were extracted from tumor segments, bounding boxes, and whole images using either classic radiomics, autoencoder, or both. Feature selection was performed to generate radiomics signatures, which were used to train machine learning algorithms for tumor categorization. Models were evaluated using the area under the receiver operating characteristic curve (AUC), sensitivity, and specificity and were statistically compared with histopathologically or follow-up-confirmed diagnosis. Results The model was developed on 1191 (mean age, 61 years ± 14 [SD]) female patients and externally validated on 50 (mean age, 55 years ± 15]). The development data set was divided into two parts: testing and training lesion segmentation (419 and 179 examinations) and lesion categorization (503 and 90 examinations). nnU-Net demonstrated precision and reproducibility in lesion segmentation in test set of data set 1 (median Dice score [DS]: 0.90 [IQR, 0.84-0.93]; = .01) and data set 2 (median DS: 0.89 [IQR, 0.80-0.92]; = .001). The best model, trained with 23 mixed features from tumor bounding boxes, achieved an AUC of 0.90 (95% CI: 0.83, 0.97), sensitivity of 81% (46 of 57; 95% CI: 70, 91), and specificity of 87% (39 of 45; 95% CI: 77, 87). No evidence of difference was found between model and human readers (AUC = 0.90 [95% CI: 0.83, 0.97] vs 0.83 [95% CI: 0.76, 0.90]; = .55 and 0.90 vs 0.82 [95% CI: 0.75, 0.90]; = .45) in tumor classification or between model and histopathologically or follow-up-confirmed diagnosis (AUC = 0.90 [95% CI: 0.83, 0.97] vs 1.00 [95% CI: 1.00,1.00]; = .10). Conclusion Precise real-time US-based breast tumor categorization was developed by mixing classic radiomics and autoencoder-based features from tumor bounding boxes. ClinicalTrials.gov identifier: NCT04976257 Published under a CC BY 4.0 license. See also the editorial by Bahl in this issue.
US 在美国临床上用于乳房成像，但其诊断性能取决于操作者的经验。计算机辅助（实时）图像分析可能有助于克服这一限制。目的：通过结合自动定位病变的经典放射组学和基于自动编码器的特征，开发基于精确实时的 US 乳房肿瘤分类方法。材料与方法：回顾性分析 2018 年 4 月至 2024 年 1 月期间的 1619 例乳腺肿瘤的 B 型超声图像。nnU-Net 用于进行病变分割。从肿瘤节段、边界框和全图像中使用经典放射组学、自动编码器或两者提取特征。进行特征选择以生成放射组学特征，这些特征用于训练用于肿瘤分类的机器学习算法。使用受试者工作特征曲线下的面积（AUC）、敏感性和特异性来评估模型，并与组织病理学或随访证实的诊断进行统计学比较。结果：该模型是在 1191 名（平均年龄 61 岁±14[SD]）女性患者中开发的，并在 50 名（平均年龄 55 岁±15）患者中进行了外部验证。开发数据集分为两部分：测试和训练病变分割（419 次和 179 次检查）和病变分类（503 次和 90 次检查）。nnU-Net 在数据集 1（中位 Dice 评分[DS]：0.90[IQR，0.84-0.93]； =.01）和数据集 2（中位 DS：0.89[IQR，0.80-0.92]； =.001）的测试数据集中具有较高的精度和可重复性。使用肿瘤边界框的 23 个混合特征训练的最佳模型获得了 0.90（95%CI：0.83，0.97）的 AUC，81%（46/57；95%CI：70，91）的敏感性和 87%（39/45；95%CI：77，87）的特异性。模型与人类读者之间在肿瘤分类（AUC = 0.90 [95%CI：0.83,0.97] vs 0.83 [95%CI：0.76,0.90]; =.55 和 0.90 vs 0.82 [95%CI：0.75,0.90]; =.45）或模型与组织病理学或随访证实的诊断（AUC = 0.90 [95%CI：0.83,0.97] vs 1.00 [95%CI：1.00,1.00]; =.10）之间无差异。结论：通过混合肿瘤边界框的经典放射组学和基于自动编码器的特征，开发了精确的实时基于 US 的乳房肿瘤分类方法。临床试验编号：NCT04976257 以 CC BY 4.0 许可发布。 另请参见本期的社论。
