Keck School of Medicine, University of Southern California, Los Angeles, CA, 90033, USA.
Department of Integrative Anatomical Sciences, University of Southern California, Los Angeles, CA, 90033, USA.
Eur Radiol. 2021 Nov;31(11):8522-8535. doi: 10.1007/s00330-021-07914-w. Epub 2021 Apr 23.
Our purpose was to differentiate between malignant from benign soft tissue neoplasms using a combination of MRI-based radiomics metrics and machine learning.
Our retrospective study identified 128 histologically diagnosed benign (n = 36) and malignant (n = 92) soft tissue lesions. 3D ROIs were manually drawn on 1 sequence of interest and co-registered to other sequences obtained during the same study. One thousand seven hundred eight radiomics features were extracted from each ROI. Univariate analyses with supportive ROC analyses were conducted to evaluate the discriminative power of predictive models constructed using Real Adaptive Boosting (Adaboost) and Random Forest (RF) machine learning approaches.
Univariate analyses demonstrated that 36.89% of individual radiomics varied significantly between benign and malignant lesions at the p ≤ 0.05 level. Adaboost and RF performed similarly well, with AUCs of 0.77 (95% CI 0.68-0.85) and 0.72 (95% CI 0.63-0.81), respectively, after 10-fold cross-validation. Restricting the machine learning models to only sequences extracted from T2FS and STIR sequences maintained comparable performance, with AUCs of 0.73 (95% CI 0.64-0.82) and 0.75 (95% CI 0.65-0.84), respectively.
Machine learning decision classifiers constructed from MRI-based radiomics features show promising ability to preoperatively discriminate between benign and malignant soft tissue masses. Our approach maintains applicability even when the dataset is restricted to T2FS and STIR fluid-sensitive sequences, which may bolster practicality in clinical application scenarios by eliminating the need for complex co-registrations for multisequence analysis.
• Predictive models constructed from MRI-based radiomics data and machine learning-augmented approaches yielded good discriminative power to correctly classify benign and malignant lesions on preoperative scans, with AUCs of 0.77 (95% CI 0.68-0.85) and 0.72 (95% CI 0.63-0.81) for Real Adaptive Boosting (Adaboost) and Random Forest (RF), respectively. • Restricting the models to only use metrics extracted from T2 fat-saturated (T2FS) and Short-Tau Inversion Recovery (STIR) sequences yielded similar performance, with AUCs of 0.73 (95% CI 0.64-0.82) and 0.75 (95% CI 0.65-0.84) for Adaboost and RF, respectively. • Radiomics-based machine learning decision classifiers constructed from multicentric data more closely mimic the real-world practice environment and warrant additional validation ahead of prospective implementation into clinical workflows.
目的：我们旨在使用基于 MRI 的放射组学指标和机器学习来区分良恶性软组织肿瘤。
方法：我们的回顾性研究纳入了 128 例经组织学诊断为良性（n=36）和恶性（n=92）的软组织病变。在感兴趣的 1 个序列上手动绘制 3D ROI，并与同一研究中获得的其他序列配准。从每个 ROI 中提取 1708 个放射组学特征。进行单变量分析和支持性 ROC 分析，以评估使用真实自适应增强（Adaboost）和随机森林（RF）机器学习方法构建的预测模型的判别能力。
结果：单变量分析表明，在 p≤0.05 水平，有 36.89%的个体放射组学特征在良性和恶性病变之间存在显著差异。Adaboost 和 RF 的表现相似，在 10 倍交叉验证后，AUC 分别为 0.77（95%CI 0.68-0.85）和 0.72（95%CI 0.63-0.81）。将机器学习模型限制为仅从 T2FS 和 STIR 序列提取的序列，可以保持类似的性能，AUC 分别为 0.73（95%CI 0.64-0.82）和 0.75（95%CI 0.65-0.84）。
结论：基于 MRI 的放射组学特征构建的机器学习决策分类器具有术前区分良恶性软组织肿块的良好能力。即使在数据集仅限于 T2FS 和 STIR 液体敏感序列的情况下，我们的方法仍然具有适用性，这可能通过消除对多序列分析的复杂配准的需求来提高临床应用场景的实用性。
关键点：
基于 MRI 的放射组学数据和机器学习增强方法构建的预测模型在术前扫描中具有良好的区分能力，可以正确分类良性和恶性病变，Adaboost 和 RF 的 AUC 分别为 0.77（95%CI 0.68-0.85）和 0.72（95%CI 0.63-0.81）。
将模型限制为仅使用从 T2 脂肪饱和（T2FS）和短回波时间反转恢复（STIR）序列提取的指标，可以得到类似的性能，Adaboost 和 RF 的 AUC 分别为 0.73（95%CI 0.64-0.82）和 0.75（95%CI 0.65-0.84）。
从多中心数据构建的基于放射组学的机器学习决策分类器更能模拟真实世界的实践环境，需要在前瞻性纳入临床工作流程之前进行更多验证。
