Department of Radiation Oncology, Sun Yat-sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangdong Key Laboratory of Nasopharyngeal Carcinoma Diagnosis and Therapy, Guangzhou, Guangdong, 510060, China.
School of Electronic and Information Engineering, South China University of Technology, Guangzhou, Guangdong, 510641, China.
Med Phys. 2021 Dec;48(12):7930-7945. doi: 10.1002/mp.15285. Epub 2021 Oct 29.
To create a network which fully utilizes multi-sequence MRI and compares favorably with manual human contouring.
We retrospectively collected 89 MRI studies of the pelvic cavity from patients with prostate cancer and cervical cancer. The dataset contained 89 samples from 87 patients with a total of 84 valid samples. MRI was performed with T1-weighted (T1), T2-weighted (T2), and Enhanced Dixon T1-weighted (T1DIXONC) sequences. There were two cohorts. The training cohort contained 55 samples and the testing cohort contained 29 samples. The MRI images in the training cohort contained contouring data from radiotherapist α. The MRI images in the testing cohort contained contouring data from radiotherapist α and contouring data from another radiotherapist: radiotherapist β. The training cohort was used to optimize the convolution neural networks, which included the attention mechanism through the proposed activation module and the blended module into multiple MRI sequences, to perform autodelineation. The testing cohort was used to assess the networks' autodelineation performance. The contoured organs at risk (OAR) were the anal canal, bladder, rectum, femoral head (L), and femoral head (R).
We compared our proposed network with UNet and FuseUNet using our dataset. When T1 was the main sequence, we input three sequences to segment five organs and evaluated the results using four metrics: the DSC (Dice similarity coefficient), the JSC (Jaccard similarity coefficient), the ASD (average mean distance), and the 95% HD (robust Hausdorff distance). The proposed network achieved improved results compared with the baselines among all metrics. The DSC were 0.834±0.029, 0.818±0.037, and 0.808±0.050 for our proposed network, FuseUNet, and UNet, respectively. The 95% HD were 7.256±2.748 mm, 8.404±3.297 mm, and 8.951±4.798 mm for our proposed network, FuseUNet, and UNet, respectively. Our proposed network also had superior performance on the JSC and ASD coefficients.
Our proposed activation module and blended module significantly improved the performance of FuseUNet for multi-sequence MRI segmentation. Our proposed network integrated multiple MRI sequences efficiently and autosegmented OAR rapidly and accurately. We also discovered that three-sequence fusion (T1-T1DIXONC-T2) was superior to two-sequence fusion (T1-T2 and T1-T1DIXONC, respectively). We infer that the more MRI sequences fused, the better the automatic segmentation results.
创建一个充分利用多序列 MRI 并优于手动轮廓勾画的网络。
我们回顾性地收集了 89 例来自前列腺癌和宫颈癌患者的盆腔 MRI 研究。该数据集包含 87 名患者的 89 个样本，共有 84 个有效样本。MRI 采用 T1 加权（T1）、T2 加权（T2）和增强 Dixon T1 加权（T1DIXONC）序列进行。有两个队列。训练队列包含 55 个样本，测试队列包含 29 个样本。训练队列中的 MRI 图像包含放射治疗师 α 的轮廓数据。测试队列中的 MRI 图像包含放射治疗师 α 和另一位放射治疗师 β 的轮廓数据。使用训练队列来优化卷积神经网络，该网络通过所提出的激活模块和混合模块包含注意力机制，以执行自动勾画。使用测试队列评估网络的自动勾画性能。勾画的危及器官（OAR）为肛门管、膀胱、直肠、股骨头（L）和股骨头（R）。
我们使用我们的数据集将我们提出的网络与 UNet 和 FuseUNet 进行了比较。当 T1 为主序列时，我们输入三个序列来分割五个器官，并使用四个指标评估结果：DSC（Dice 相似系数）、JSC（Jaccard 相似系数）、ASD（平均平均距离）和 95% HD（稳健 Hausdorff 距离）。与基线相比，所提出的网络在所有指标中都取得了更好的结果。我们提出的网络的 DSC 分别为 0.834±0.029、0.818±0.037 和 0.808±0.050，FuseUNet 和 UNet 的 DSC 分别为 0.818±0.037 和 0.808±0.050。我们提出的网络的 95% HD 分别为 7.256±2.748mm、8.404±3.297mm 和 8.951±4.798mm，FuseUNet 和 UNet 的 95% HD 分别为 8.404±3.297mm 和 8.951±4.798mm。我们提出的网络在 JSC 和 ASD 系数上也表现出更好的性能。
我们提出的激活模块和混合模块显著提高了 FuseUNet 对多序列 MRI 分割的性能。我们提出的网络有效地集成了多个 MRI 序列，并快速准确地自动勾画 OAR。我们还发现，三序列融合（T1-T1DIXONC-T2）优于两序列融合（T1-T2 和 T1-T1DIXONC）。我们推断，融合的 MRI 序列越多，自动分割的结果越好。
