KneePreM：通过大规模无标签预训练和标签高效微调迈向3D膝关节MRI基础模型
KneePreM: Towards 3D Knee MRI Foundation Models via Large-Scale Unlabeled Pretraining and Label-Efficient Fine-Tuning
- Cleveland Clinic（克利夫兰诊所）
- Lerner Research Institute（勒纳研究所）
- Case Western Reserve University（凯斯西储大学）
- IBM Research（IBM研究院）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
KneePreM通过大规模无标签预训练和标签高效微调，提升了膝关节MRI分类与分割的迁移性能和标签效率，尤其在数据有限时优于基线。
AI中文摘要：
背景：各数据仓库中存有大量无标签的膝关节MRI扫描数据，但这些数据尚未得到充分利用。我们开发了KneePreM，一种膝关节专用的3D自监督模型，并评估了其在分类和分割任务中的迁移能力和标签效率。方法：我们在来自4,791名参与者的19,011个无标签骨关节炎倡议（OAI）MRI序列上预训练了一个3D U-Net掩码自编码器。下游微调使用了完整和缩减的训练集，用于fastMRI+二标签分类（1,172次检查）、关节镜部分半月板切除术（APM）八目标分类（1,716次检查）、SKM-TEA分割（155次检查）和APM分割（25次检查）。基线为随机初始化和SuPreM。部署工作流通过模型上下文协议接口实现。评估指标包括平衡准确率、F1分数、ROC AUC、PR AUC和Dice分数。统计分析使用自助法置信区间和配对自助检验进行分类，使用Wilcoxon符号秩检验进行分割。结果：KneePreM在fastMRI+和APM上均实现了比两个基线更高的全数据宏ROC AUC（所有p < .001）。对于fastMRI+分类，KneePreM使用50%的训练数据达到了0.722的ROC AUC，超过了两个全数据基线。在APM分类中，KneePreM使用70%的数据达到了0.740的ROC AUC，匹配了全数据随机基线并优于SuPreM。对于SKM-TEA分割，其70%数据的Dice分数为0.838，超过了全数据随机基线（0.835）和两个同预算比较器。在APM分割中，其75%数据的Dice分数为0.746，超过了全数据随机基线（0.731）和两个同预算比较器。结论：KneePreM在膝关节MRI分类和分割任务中提高了迁移性能和标签效率，尤其是在标记训练数据有限的情况下。
英文摘要：
Background: Large volumes of unlabeled knee MRI scans are available across repositories but remain insufficiently leveraged. We developed KneePreM, a knee-specific 3D self-supervised model, and evaluated transfer and label efficiency for classification and segmentation. Methods: A 3D U-Net masked autoencoder was pretrained on 19,011 unlabeled Osteoarthritis Initiative (OAI) MRI series from 4,791 participants. Downstream fine-tuning used full and reduced training sets for fastMRI+ two-label classification (1,172 examinations), Arthroscopic Partial Meniscectomy (APM) eight-target classification (1,716 examinations), SKM-TEA segmentation (155 examinations), and APM segmentation (25 examinations). Baselines were random initialization and SuPreM. Deployment workflow was implemented with a Model Context Protocol interface. Evaluation metrics included balanced accuracy, F1 score, ROC AUC, PR AUC, and Dice score. Statistical analysis used bootstrap confidence intervals and paired bootstrap tests for classification and Wilcoxon signed-rank tests for segmentation. Results: KneePreM achieved higher full-data macro ROC AUC than both baselines for fastMRI+ and APM (all p < .001). For fastMRI+ classification, KneePreM achieved a ROC AUC of 0.722 using 50% of the training data, exceeding both full-data baselines. In APM classification, KneePreM reached a ROC AUC of 0.740 with 70% of the data, matching the full-data random baseline and outperforming SuPreM. For SKM-TEA segmentation, its 70%-data Dice of 0.838 exceeded the full-data random baseline (0.835) and both same-budget comparators. In APM segmentation, its 75%-data Dice of 0.746 exceeded the full-data random baseline (0.731) and both same-budget comparators. Conclusion: KneePreM improves transfer performance and label efficiency across knee MRI classification and segmentation tasks, particularly when labeled training data are limited.
