Department of Radiology, Systems Medicine and Bioengineering, Houston Methodist Cancer Center, Houston Methodist Hospital, Weill Cornell Medicine, TX (K.K.W., J.S.C., R.G., S.T.C.W.).
The Ting Tsung and Wei Fong Chao Center for BRAIN, Houston Methodist Hospital, TX (K.K.W., S.T.C.W.).
Stroke. 2022 Sep;53(9):2896-2905. doi: 10.1161/STROKEAHA.121.037982. Epub 2022 May 12.
Stroke infarct volume predicts patient disability and has utility for clinical trial outcomes. Accurate infarct volume measurement requires manual segmentation of stroke boundaries in diffusion-weighted magnetic resonance imaging scans which is time-consuming and subject to variability. Automatic infarct segmentation should be robust to rotation and reflection; however, prior work has not encoded this property into deep learning architecture. Here, we use rotation-reflection equivariance and train a deep learning model to segment stroke volumes in a large cohort of well-characterized patients with acute ischemic stroke in different vascular territories.
In this retrospective study, patients were selected from a stroke registry at Houston Methodist Hospital. Eight hundred seventy-five patients with acute ischemic stroke in any brain area who had magnetic resonance imaging with diffusion-weighted imaging were included for analysis and split 80/20 for training/testing. Infarct volumes were manually segmented by consensus of 3 independent clinical experts and cross-referenced against radiology reports. A rotation-reflection equivariant model was developed based on U-Net and grouped convolutions. Segmentation performance was evaluated using Dice score, precision, and recall. Ninety-day modified Rankin Scale outcome prediction was also evaluated using clinical variables and segmented stroke volumes in different brain regions.
Segmentation model Dice scores are 0.88 (95% CI, 0.87-0.89; training) and 0.85 (0.82-0.88; testing). The modified Rankin Scale outcome prediction AUC using stroke volume in 30 refined brain regions based upon modified Rankin Scale-relevance areas adjusted for clinical variables was 0.80 (0.76-0.83) with an accuracy of 0.75 (0.72-0.78).
We trained a deep learning model with encoded rotation-reflection equivariance to segment acute ischemic stroke lesions in diffusion- weighted imaging using a large data set from the Houston Methodist stroke center. The model achieved competitive performance in 175 well-balanced hold-out testing cases that include strokes from different vascular territories. Furthermore, the location specific stroke volume segmentations from the deep learning model combined with clinical factors demonstrated high AUC and accuracy for 90-day modified Rankin Scale in an outcome prediction model.
中风梗死体积预测患者残疾程度，并对临床试验结果具有实用性。准确测量梗死体积需要在弥散加权磁共振成像扫描中手动分割中风边界，这既耗时又容易出现变化。自动梗死分割应该对旋转和反射具有鲁棒性；然而，先前的工作并没有将此属性编码到深度学习架构中。在这里，我们使用旋转反射等价性，并训练一个深度学习模型，以分割在不同血管区域具有急性缺血性中风的大量特征良好的患者的弥散加权成像中的中风体积。
在这项回顾性研究中，患者是从休斯顿卫理公会医院的中风登记处中选择的。共有 875 名患有急性缺血性中风的患者，其任何脑区均进行了磁共振成像弥散加权成像，他们被纳入分析，并按 80/20 进行训练/测试。梗死体积由 3 位独立临床专家的共识手动分割，并与放射学报告交叉引用。基于 U-Net 和分组卷积开发了一个旋转反射等价模型。使用 Dice 分数、精度和召回率评估分割性能。使用临床变量和不同脑区的分割中风体积也评估了 90 天改良 Rankin 量表预后预测。
分割模型的 Dice 分数分别为 0.88（95%置信区间，0.87-0.89；训练）和 0.85（0.82-0.88；测试）。使用基于改良 Rankin 量表相关性区域并根据临床变量调整的 30 个细化脑区的中风体积进行的改良 Rankin 量表预后预测 AUC 为 0.80（0.76-0.83），准确性为 0.75（0.72-0.78）。
我们使用来自休斯顿卫理公会中风中心的大型数据集训练了一个具有旋转反射等价编码的深度学习模型，以分割弥散加权成像中的急性缺血性中风病变。该模型在 175 个平衡的测试案例中表现出了有竞争力的性能，其中包括来自不同血管区域的中风。此外，深度学习模型的特定位置中风体积分割与临床因素相结合，在预后模型中表现出了 90 天改良 Rankin 量表的高 AUC 和准确性。
