[论文解读] White matter hyperintensity and stroke lesion segmentation and differentiation using convolutional neural networks

Ricardo Guerrero, Chen Qin+9 more|arXiv (Cornell University)|Jun 3, 2017

Brain Tumor Detection and Classification参考文献 59被引用 18

一句话总结

本文提出uResNet，一种全卷积神经网络，可同时在FLAIR MRI扫描中分割白质高信号（WMH）并将其与脑卒中病灶区分开来。该模型在WMH上的平均Dice分数达到66.9±18.1，在脑卒中病灶上的平均Dice分数达到40.0±25.2，性能优于现有方法，并与专家评分及临床结局具有强相关性。

ABSTRACT

The accurate assessment of White matter hyperintensities (WMH) burden is of crucial importance for epidemiological studies to determine association between WMHs, cognitive and clinical data. The manual delineation of WMHs is tedious, costly and time consuming. This is further complicated by the fact that other pathological features (i.e. stroke lesions) often also appear as hyperintense. Several automated methods aiming to tackle the challenges of WMH segmentation have been proposed, however cannot differentiate between WMH and strokes. Other methods, capable of distinguishing between different pathologies in brain MRI, are not designed with simultaneous WMH and stroke segmentation in mind. In this work we propose to use a convolutional neural network (CNN) that is able to segment hyperintensities and differentiate between WMHs and stroke lesions. Specifically, we aim to distinguish between WMH pathologies from those caused by stroke lesions due to either cortical, large or small subcortical infarcts. As far as we know, this is the first time such differentiation task has explicitly been proposed. The proposed fully convolutional CNN architecture, is comprised of an analysis path, that gradually learns low and high level features, followed by a synthesis path, that gradually combines and up-samples the low and high level features into a class likelihood semantic segmentation. Quantitatively, the proposed CNN architecture is shown to outperform other well established and state-of-the-art algorithms in terms of overlap with manual expert annotations. Clinically, the extracted WMH volumes were found to correlate better with the Fazekas visual rating score. Additionally, a comparison of the associations found between clinical risk-factors and the WMH volumes generated by the proposed method, were found to be in line with the associations found with the expert-annotated volumes.

研究动机与目标

• 开发一种完全自动化、任务特定的方法，用于在FLAIR MRI扫描中分割并区分白质高信号（WMH）与脑卒中病灶。
• 解决现有方法的临床局限性，即无法区分WMH与脑卒中病灶，而两者病理机制与治疗意义不同。
• 提高WMH负荷量化在流行病学研究、临床试验和纵向评估中的准确性与可靠性。
• 与竞争性自动化方法相比，实现与Fazekas视觉评分及临床危险因素更强的相关性。
• 提供一种临床可用的解决方案，减少对放射科专家耗时的手动分割的依赖。

提出的方法

• 所提方法采用一种类似U-Net的全卷积神经网络结构，称为uResNet，结合编码器路径进行特征提取，以及解码器路径进行特征融合与上采样。
• 网络通过多类别交叉熵损失函数进行端到端训练，以预测健康组织、WMH和脑卒中病灶的体素级分类概率。
• 训练过程中应用基于图像块偏移（非中心采样）的数据增强技术，以提高模型鲁棒性与泛化能力，尤其针对小尺寸或不规则形状的病灶。
• 模型在多对比度MRI数据上进行训练，包括FLAIR、T1加权像、白质概率图以及脑脊液（CSF）图谱，以提升病灶定位精度。
• 全卷积设计支持在整个3D体积上进行密集预测，无需分块推理，确保空间一致性。
• 编码器部分采用残差块（ResNet风格），以稳定训练过程并改善深层网络中的梯度流动。

实验结果

研究问题

• RQ1 深度学习模型能否在FLAIR MRI扫描中有效区分白质高信号与脑卒中病灶？
• RQ2 所提出的uResNet模型在WMH与脑卒中病灶分割任务中是否显著优于现有最先进方法？
• RQ3 uResNet自动分割的WMH体积与专家标注体积及Fazekas视觉评分的相关性如何？
• RQ4 自动化WMH体积与已知临床危险因素（如年龄、高血压、认知功能）的关联性如何？
• RQ5 训练过程中使用非中心图像块采样是否能提升模型在小尺寸或不规则病灶上的泛化能力与性能？

主要发现

• uResNet模型在WMH上的平均Dice分数为66.9±18.1，在脑卒中病灶上的平均Dice分数为40.0±25.2，显著优于基线方法（包括DeepMedic及其他最先进模型）。
• 自动化WMH体积与Fazekas视觉评分之间的相关性（皮尔逊相关系数CC = 0.824）高于其他竞争方法，且与专家标注体积的相关性相当。
• 模型展现出强临床有效性，其与WMH体积相关的危险因素（如年龄、高血压）关联性与专家标注结果高度一致。
• 采用非中心图像块采样显著提升了性能，Dice分数较常规采样有统计学显著提高（p<0.001）。
• 增加多模态输入（如T1、CSF图谱、白质图）可提升性能，但增益有限且带来额外计算开销，表明仅使用FLAIR图像已足以实现稳健分割。
• 尽管进行了广泛超参数调优，DeepMedic在相同训练设置下仍无法收敛或获得稳定结果，表明uResNet具有更优的训练稳定性。