指南与共识神经MR

多发性硬化脑MRI研究病灶分割算法共识

Consensus of algorithms for lesion segmentation in brain MRI studies of multiple sclerosis.
指南 / 共识 · Sci Rep · 2024-09-12

本研究提出一种基于共识的框架,通过组合多个公开的先进深度学习模型对T1加权和FLAIR扫描进行多发性硬化病灶分割,并在两个中心(私人数据集131例、公共数据集30例)的MRI数据上评估;私人数据集Dice提高4-8%、精确率提高3-4%,公共数据集Dice提高4-8%、精确率提高3-4%,估计病灶负荷与真实值高度一致(ρ=0.92和ρ=0.97),病灶计数一致性ρ=0.83和ρ=0.94。

要点

  • 共识框架组合多个公开深度学习模型,无需额外训练即可提升病灶分割重叠度和检测分数。
  • 私人数据集Dice提高4-8%、精确率提高3-4%;公共数据集Dice提高4-8%、精确率提高3-4%。
  • 估计病灶负荷与真实值高度一致(私人ρ=0.92、公共ρ=0.97),病灶计数一致性私人ρ=0.83、公共ρ=0.94。

核心结论

  • 该共识框架利用互补特征并克服单个算法的一些局限性,确保准确可靠的分割结果。

研究设计与样本

  • 基于共识的框架,组合公开的先进深度学习模型,运行多个分割任务后合并各算法输出,未对任何纳入算法进行进一步训练;评估使用两个中心的MRI数据集,私人数据集131例多发性硬化患者、公共数据集30例多发性硬化患者,均具有手动分割的病灶掩膜。

诊断标准与关键数值

  • 私人数据集Dice提高4-8%,精确率提高3-4%。
  • 公共数据集Dice提高4-8%,精确率提高3-4%。
  • 私人数据集估计病灶负荷与真实值一致性ρ=0.92,病灶计数一致性ρ=0.83。
  • 公共数据集估计病灶负荷与真实值一致性ρ=0.97,病灶计数一致性ρ=0.94。

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

Segmentation of multiple sclerosis (MS) lesions on brain MRI scans is crucial for diagnosis, disease and treatment monitoring but is a time-consuming task. Despite several automated algorithms have been proposed, there is still no consensus on the most effective method. Here, we applied a consensus-based framework to improve lesion segmentation on T1-weighted and FLAIR scans. The framework is designed to combine publicly available state-of-the-art deep learning models, by running multiple segmentation tasks before merging the outputs of each algorithm. To assess the effectiveness of the approach, we applied it to MRI datasets from two different centers, including a private and a public dataset, with 131 and 30 MS patients respectively, with manually segmented lesion masks available. No further training was performed for any of the included algorithms. Overlap and detection scores were improved, with Dice increasing by 4-8% and precision by 3-4% respectively for the private and public dataset. High agreement was obtained between estimated and true lesion load (ρ = 0.92 and ρ = 0.97) and count (ρ = 0.83 and ρ = 0.94). Overall, this framework ensures accurate and reliable results, exploiting complementary features and overcoming some of the limitations of individual algorithms.

来源:Sci Rep,查看原文。版权归原作者所有。

引用

  1. De Rosa AP, Benedetto M, Tagliaferri S, et al. Consensus of algorithms for lesion segmentation in brain MRI studies of multiple sclerosis. Sci Rep. 2024. PMID: 39266642. doi:10.1038/s41598-024-72649-9

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。