Automatic segmentation of rectal tumor on diffusion-weighted images by deep learning with U-Net.
Automatic segmentation of rectal tumor on diffusion-weighted images by deep learning with U-Net.
复制标题
通过 U-Net 深度学习在扩散加权图像上自动分割直肠肿瘤
DOI：
10.1002/acm2.13381
复制
发表时间：
2021-09
影响因子：
2.1
通讯作者：
Sun YS
中科院分区：
文献类型：
--
作者：
Zhu HT;Zhang XY;Shi YJ;Li XT;Sun YS
在容积图像上手动描绘直肠肿瘤是耗时且主观的。深度学习已被用于在T2加权图像上自动分割直肠肿瘤，但弥散加权成像的自动分割受到噪声、伪影和低分辨率的挑战。在这项研究中，提出了一种体积U形神经网络（U-Net），用于在扩散加权图像上自动分割直肠肿瘤。300例局部晚期直肠癌患者参加了这项研究，并分为培训组，验证组和测试组。由经验丰富的放射科医生在弥散加权图像上描绘直肠肿瘤区域作为基础事实。设计了一个U‐Net，其体积输入为扩散加权图像，输出为相同大小的分割。使用半自动分割方法进行比较，通过手动选择灰度阈值并自动选择最大连通区域。通过计算Dice相似系数（DSC）对方法进行评价。在测试组中，深度学习方法（DSC = 0.675 ± 0.144，中位DSC为0.702，最大DSC为0.893，最小DSC为0.297）显示出比半自动方法（DSC = 0.614 ± 0.225，中位DSC为0.685，最大DSC为0.869，最小DSC为0.047）更高的分割准确性。配对t检验显示，测试组中深度学习方法和半自动方法之间的DSC存在显著差异（T = 2.160，p = 0.035）。Volumetric U-Net可以在局部晚期直肠癌的DWI图像上自动分割直肠肿瘤区域。
Manual delineation of a rectal tumor on a volumetric image is time‐consuming and subjective. Deep learning has been used to segment rectal tumors automatically on T2‐weighted images, but automatic segmentation on diffusion‐weighted imaging is challenged by noise, artifact, and low resolution. In this study, a volumetric U‐shaped neural network (U‐Net) is proposed to automatically segment rectal tumors on diffusion‐weighted images. Three hundred patients of locally advanced rectal cancer were enrolled in this study and divided into a training group, a validation group, and a test group. The region of rectal tumor was delineated on the diffusion‐weighted images by experienced radiologists as the ground truth. A U‐Net was designed with a volumetric input of the diffusion‐weighted images and an output of segmentation with the same size. A semi‐automatic segmentation method was used for comparison by manually choosing a threshold of gray level and automatically selecting the largest connected region. Dice similarity coefficient (DSC) was calculated to evaluate the methods. On the test group, deep learning method (DSC = 0.675 ± 0.144, median DSC is 0.702, maximum DSC is 0.893, and minimum DSC is 0.297) showed higher segmentation accuracy than the semi‐automatic method (DSC = 0.614 ± 0.225, median DSC is 0.685, maximum DSC is 0.869, and minimum DSC is 0.047). Paired t‐test shows significant difference (T = 2.160, p = 0.035) in DSC between the deep learning method and the semi‐automatic method in the test group. Volumetric U‐Net can automatically segment rectal tumor region on DWI images of locally advanced rectal cancer.
登录
查看更多内容
影响因子：
19.7
作者：
Horvat, Natally;Veeraraghavan, Harini;Petkovska, Iva
通讯作者：
Petkovska, Iva
DOI：
10.6004/jnccn.2018.0061
发表时间：
2018-07
期刊：
Journal of the National Comprehensive Cancer Network : JNCCN
影响因子：
--
作者：
Benson AB;Venook AP;Al-Hawary MM;Cederquist L;Chen YJ;Ciombor KK;Cohen S;Cooper HS;Deming D;Engstrom PF;Grem JL;Grothey A;Hochster HS;Hoffe S;Hunt S;Kamel A;Kirilcuk N;Krishnamurthi S;Messersmith WA;Meyerhardt J;Mulcahy MF;Murphy JD;Nurkin S;Saltz L;Sharma S;Shibata D;Skibber JM;Sofocleous CT;Stoffel EM;Stotsky-Himelfarb E;Willett CG;Wuthrick E;Gregory KM;Gurski L;Freedman-Cass DA
通讯作者：
Freedman-Cass DA
影响因子：
3.8
作者：
Liu, Xiaoming;Guo, Shuxu;Li, Xueyan
通讯作者：
Li, Xueyan
扩散加权成像的定量分析预测局部晚期直肠癌新辅助放化疗的病理良好反应
影响因子：
5.7
作者：
Tang, Zhenchao;Zhang, Xiao-Yan;Tian, Jie
通讯作者：
Tian, Jie
影响因子：
19.7
作者：
Sun, Ying-Shi;Zhang, Xiao-Peng;Zhang, Xiao-Yan
通讯作者：
Zhang, Xiao-Yan