Department of Convergence Medicine, Biomedical Engineering Research Center, University of Ulsan College of Medicine, Asan Medical Center, Seoul, Korea.
Department of Radiology and Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, Korea.
Korean J Radiol. 2019 Aug;20(8):1275-1284. doi: 10.3348/kjr.2018.0615.
To develop algorithms using convolutional neural networks (CNNs) for automatic segmentation of acute ischemic lesions on diffusion-weighted imaging (DWI) and compare them with conventional algorithms, including a thresholding-based segmentation.
Between September 2005 and August 2015, 429 patients presenting with acute cerebral ischemia (training:validation:test set = 246:89:94) were retrospectively enrolled in this study, which was performed under Institutional Review Board approval. Ground truth segmentations for acute ischemic lesions on DWI were manually drawn under the consensus of two expert radiologists. CNN algorithms were developed using two-dimensional U-Net with squeeze-and-excitation blocks (U-Net) and a DenseNet with squeeze-and-excitation blocks (DenseNet) with squeeze-and-excitation operations for automatic segmentation of acute ischemic lesions on DWI. The CNN algorithms were compared with conventional algorithms based on DWI and the apparent diffusion coefficient (ADC) signal intensity. The performances of the algorithms were assessed using the Dice index with 5-fold cross-validation. The Dice indices were analyzed according to infarct volumes (< 10 mL, ≥ 10 mL), number of infarcts (≤ 5, 6-10, ≥ 11), and b-value of 1000 (b1000) signal intensities (< 50, 50-100, > 100), time intervals to DWI, and DWI protocols.
The CNN algorithms were significantly superior to conventional algorithms ( < 0.001). Dice indices for the CNN algorithms were 0.85 for U-Net and DenseNet and 0.86 for an ensemble of U-Net and DenseNet, while the indices were 0.58 for ADC-b1000 and b1000-ADC and 0.52 for the commercial ADC algorithm. The Dice indices for small and large lesions, respectively, were 0.81 and 0.88 with U-Net, 0.80 and 0.88 with DenseNet, and 0.82 and 0.89 with the ensemble of U-Net and DenseNet. The CNN algorithms showed significant differences in Dice indices according to infarct volumes ( < 0.001).
The CNN algorithm for automatic segmentation of acute ischemic lesions on DWI achieved Dice indices greater than or equal to 0.85 and showed superior performance to conventional algorithms.
开发基于卷积神经网络（CNN）的算法，用于对弥散加权成像（DWI）上的急性缺血性病变进行自动分割，并与传统算法进行比较，包括基于阈值的分割。
本研究回顾性纳入了 2005 年 9 月至 2015 年 8 月期间 429 例急性脑缺血患者（训练：验证：测试集=246：89：94），本研究获得了机构审查委员会的批准。两名专家放射科医生根据共识对 DWI 上的急性缺血性病变进行了手动分割。使用二维 U-Net 与挤压激励块（U-Net）和 DenseNet 与挤压激励块（DenseNet）开发了 CNN 算法，用于对 DWI 上的急性缺血性病变进行自动分割。基于 DWI 和表观弥散系数（ADC）信号强度对 CNN 算法和传统算法进行了比较。使用 5 折交叉验证评估算法的性能，并通过 Dice 指数进行分析。根据梗死体积（<10 mL，≥10 mL）、梗死数量（≤5、6-10、≥11）和 b 值为 1000 的信号强度（<50、50-100、>100）、DWI 时间间隔和 DWI 方案对 Dice 指数进行了分析。
CNN 算法明显优于传统算法（<0.001）。U-Net 和 DenseNet 的 Dice 指数分别为 0.85 和 0.86，U-Net 和 DenseNet 联合的 Dice 指数为 0.86，而 ADC-b1000 和 b1000-ADC 的指数分别为 0.58 和 0.52，商用 ADC 算法的指数为 0.52。U-Net 的小病变和大病变的 Dice 指数分别为 0.81 和 0.88，DenseNet 分别为 0.80 和 0.88，U-Net 和 DenseNet 联合的分别为 0.82 和 0.89。CNN 算法根据梗死体积显示出 Dice 指数的显著差异（<0.001）。
用于 DWI 上急性缺血性病变自动分割的 CNN 算法获得了大于或等于 0.85 的 Dice 指数，并表现出优于传统算法的性能。
