文献速递神经MR

医学影像基础模型辅助脑肿瘤诊断:一项初步研究

Medical image foundation models in assisting diagnosis of brain tumors: a pilot study.
PubMed 文献 · Eur Radiol · 2024-04-16

本回顾性研究利用57,621例增强头颅MRI扫描,通过自监督学习构建全脑多对比度MRI基础模型,并在下游任务中微调用于脑肿瘤检测、鉴别及分子状态预测。最佳分类器在独立测试集上的准确率分别为94.9%、92.3%和80.4%,对应AUC为0.981、0.972和0.852;预训练权重分类器在所有任务上准确率和AUC均优于从头训练的卷积分类器约10%。

要点

  • 自监督基础模型可从多对比度全脑MRI中提取有效表征,用于脑肿瘤检测、鉴别和分子状态预测。
  • 最佳分类器在独立测试集上检测、鉴别和分子状态预测的准确率分别为94.9%、92.3%和80.4%。
  • 预训练权重分类器在所有任务上准确率和AUC均优于从头训练的卷积分类器约10%。

核心结论

  • 自监督学习获得的基础模型在脑肿瘤下游任务中展现出可扩展性和可解释性的潜力,并有望扩展至病灶弥漫分布的神经系统疾病。

研究设计与样本

  • 回顾性研究,使用57,621例增强头颅MRI扫描,通过自监督学习构建基础模型,预训练任务为跨对比度上下文恢复,采用两种内容丢弃方案;下游分类器基于预训练基础模型构建并微调,用于脑肿瘤检测、鉴别和分子状态预测;以仅在 downstream 任务数据上训练的卷积神经网络作为对照。

关键影像表现/征象

  • 正确预测病例的显著区域主要聚集在肿瘤周围。
  • 预训练基础模型能够从多对比度全脑 volumes 中提取有效表征。

诊断标准与关键数值

  • 脑肿瘤检测:最佳分类器准确率94.9%,AUC 0.981。
  • 脑肿瘤鉴别:最佳分类器准确率92.3%,AUC 0.972。
  • 分子状态预测:最佳分类器准确率80.4%,AUC 0.852。
  • 预训练权重分类器在所有任务上准确率和AUC均优于从头训练的卷积分类器约10%。
  • 两种内容丢弃方案衍生的分类器仅在脑肿瘤检测上存在显著差异。

临床应用与阅片要点

  • 本研究方法用于预测胶质瘤关键分子状态,预期可改善治疗计划和患者结局。
  • 所开发的基础模型可作为推进脑部疾病诊断人工智能应用的基石。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

英文原文(摘要)· 点击展开

OBJECTIVES: To build self-supervised foundation models for multicontrast MRI of the whole brain and evaluate their efficacy in assisting diagnosis of brain tumors.

METHODS: In this retrospective study, foundation models were developed using 57,621 enhanced head MRI scans through self-supervised learning with a pretext task of cross-contrast context restoration with two different content dropout schemes. Downstream classifiers were constructed based on the pretrained foundation models and fine-tuned for brain tumor detection, discrimination, and molecular status prediction. Metrics including accuracy, sensitivity, specificity, and area under the ROC curve (AUC) were used to evaluate the performance. Convolutional neural networks trained exclusively on downstream task data were employed for comparative analysis.

RESULTS: The pretrained foundation models demonstrated their ability to extract effective representations from multicontrast whole-brain volumes. The best classifiers, endowed with pretrained weights, showed remarkable performance with accuracies of 94.9, 92.3, and 80.4%, and corresponding AUC values of 0.981, 0.972, and 0.852 on independent test datasets in brain tumor detection, discrimination, and molecular status prediction, respectively. The classifiers with pretrained weights outperformed the convolutional classifiers trained from scratch by approximately 10% in terms of accuracy and AUC across all tasks. The saliency regions in the correctly predicted cases are mainly clustered around the tumors. Classifiers derived from the two dropout schemes differed significantly only in the detection of brain tumors.

CONCLUSIONS: Foundation models obtained from self-supervised learning have demonstrated encouraging potential for scalability and interpretability in downstream brain tumor-related tasks and hold promise for extension to neurological diseases with diffusely distributed lesions.

CLINICAL RELEVANCE STATEMENT: The application of our proposed method to the prediction of key molecular status in gliomas is expected to improve treatment planning and patient outcomes. Additionally, the foundation model we developed could serve as a cornerstone for advancing AI applications in the diagnosis of brain-related diseases.

来源:Eur Radiol,查看原文。版权归原作者所有。

引用

  1. Chen M, Zhang M, Yin L, et al. Medical image foundation models in assisting diagnosis of brain tumors: a pilot study. Eur Radiol. 2024. PMID: 38627290. doi:10.1007/s00330-024-10728-1

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。