Evaluating Multi-Task Morphological Concept Learning for Pulmonary Nodule Malignancy Assessment in 3D CT
Evaluating Multi-Task Morphological Concept Learning for Pulmonary Nodule Malignancy Assessment in 3D CT
评估用于 3D CT 肺结节恶性程度评估的多任务形态学概念学习
Abstract: Morphological characteristics such as spiculation and lobulation play an important role in assessing pulmonary nodules on computed tomography (CT), particularly in relation to malignancy risk. This study examines whether learning radiologist-annotated morphological features together with malignancy risk from lesion-centred 3D CT volumes improves classification performance.
摘要： 形态学特征（如毛刺征和分叶征）在评估计算机断层扫描（CT）中的肺结节时起着重要作用，特别是在与恶性风险相关时。本研究探讨了从以病灶为中心的 3D CT 体数据中，同时学习放射科医生标注的形态学特征与恶性风险，是否能提高分类性能。
The Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI) dataset was used, comprising 3,918 reader-level nodule annotations from 742 patients after excluding indeterminate malignancy ratings. Patient-level splitting was used for training, validation, and testing, with 112 patients and 628 reader annotations in the held-out test set.
本研究使用了肺图像数据库联盟和图像数据库资源倡议（LIDC-IDRI）数据集，在排除恶性程度不确定的评分后，包含来自 742 名患者的 3,918 条读者级结节标注。研究采用患者级划分进行训练、验证和测试，其中留出的测试集包含 112 名患者和 628 条读者标注。
A single-task 3D convolutional neural network was compared with a multi-task model predicting malignancy risk, spiculation, and lobulation. The single-task model achieved a balanced accuracy of 0.548 and receiver operating characteristic area under the curve (ROC-AUC) of 0.552, while the multi-task model achieved 0.539 and 0.558, respectively.
研究对比了单任务 3D 卷积神经网络与预测恶性风险、毛刺征和分叶征的多任务模型。单任务模型实现了 0.548 的平衡准确率和 0.552 的受试者工作特征曲线下面积（ROC-AUC），而多任务模型分别为 0.539 和 0.558。
Patient-level bootstrap analysis showed an ROC-AUC difference of 0.005 (95% confidence interval (CI): -0.087 to 0.090) and a balanced-accuracy difference of -0.009 (95% CI: -0.067 to 0.043). The auxiliary tasks were strongly imbalanced and showed limited predictive performance.
患者级自助法（bootstrap）分析显示，ROC-AUC 的差异为 0.005（95% 置信区间 (CI)：-0.087 至 0.090），平衡准确率的差异为 -0.009（95% CI：-0.067 至 0.043）。辅助任务存在严重的类别不平衡，且表现出有限的预测性能。
Overall, including morphological features did not clearly improve malignancy-risk classification, showing the importance of class balance, label formulation, and reader-level annotation structure in multi-task pulmonary CT analysis.
总体而言，纳入形态学特征并未明显改善恶性风险的分类效果，这表明在多任务肺部 CT 分析中，类别平衡、标签制定以及读者级标注结构的重要性。