Yang Yiyuan, Tang Yucheng, Gao Riqiang, Bao Shunxing, Huo Yuankai, McKenna Matthew T, Savona Michael R, Abramson Richard G, Landman Bennett A
Vanderbilt University, Department of Electrical Engineering and Computer Science, Nashville, Tennessee, United States.
Vanderbilt University School of Medicine, Vanderbilt-Ingram Cancer Center, Nashville, Tennessee, United States.
J Med Imaging (Bellingham). 2021 Jan;8(1):014004. doi: 10.1117/1.JMI.8.1.014004. Epub 2021 Feb 19.
Deep learning is a promising technique for spleen segmentation. Our study aims to validate the reproducibility of deep learning-based spleen volume estimation by performing spleen segmentation on clinically acquired computed tomography (CT) scans from patients with myeloproliferative neoplasms. As approved by the institutional review board, we obtained 138 de-identified abdominal CT scans. A sum of voxel volume on an expert annotator's segmentations establishes the ground truth (estimation 1). We used our deep convolutional neural network (estimation 2) alongside traditional linear estimations (estimation 3 and 4) to estimate spleen volumes independently. Dice coefficient, Hausdorff distance, coefficient, Pearson coefficient, the absolute difference in volume, and the relative difference in volume were calculated for 2 to 4 against the ground truth to compare and assess methods' performances. We re-labeled on scan-rescan on a subset of 40 studies to evaluate method reproducibility. Calculated against the ground truth, the coefficients for our method (estimation 2) and linear method (estimation 3 and 4) are 0.998, 0.954, and 0.973, respectively. The Pearson coefficients for the estimations against the ground truth are 0.999, 0.963, and 0.978, respectively (paired -tests produced between 2 and 3, and 2 and 4). The deep convolutional neural network algorithm shows excellent potential in rendering more precise spleen volume estimations. Our computer-aided segmentation exhibits reasonable improvements in splenic volume estimation accuracy.
深度学习是一种很有前景的脾脏分割技术。我们的研究旨在通过对骨髓增殖性肿瘤患者的临床计算机断层扫描（CT）图像进行脾脏分割，来验证基于深度学习的脾脏体积估计的可重复性。经机构审查委员会批准，我们获取了138份去识别化的腹部CT扫描图像。专家标注的分割图像上的体素体积总和确定了真实值（估计值1）。我们使用深度卷积神经网络（估计值2）以及传统线性估计方法（估计值3和4）来独立估计脾脏体积。计算估计值2至4与真实值之间的骰子系数、豪斯多夫距离、 系数、皮尔逊系数、体积绝对差和体积相对差，以比较和评估各方法的性能。我们对40项研究的子集进行了扫描-再扫描重新标注，以评估方法的可重复性。相对于真实值计算，我们的方法（估计值2）和线性方法（估计值3和4）的 系数分别为0.998、0.954和0.973。相对于真实值的估计的皮尔逊系数分别为0.999、0.963和0.978（配对 检验得出2与3之间以及2与4之间的 ）。深度卷积神经网络算法在进行更精确的脾脏体积估计方面显示出巨大潜力。我们的计算机辅助分割在脾脏体积估计准确性方面有合理的提升。
