| 刘想,韩超,高歌,朱丽娜,陈卫东,黄嘉豪,王祥鹏,张晓东,王霄英.3D U-Net深度学习模型基于盆腔T2WI自动分割盆腔软组织结构[J].中国医学影像技术,2022,38(2):266~271 |
| 3D U-Net深度学习模型基于盆腔T2WI自动分割盆腔软组织结构 |
| 3D U-Net deep learning model for automatic segmentation of pelvic soft tissue structures based on pelvic T2WI |
| 投稿时间：2021-04-19 修订日期：2021-08-13 |
| DOI：10.13929/j.issn.1003-3289.2022.02.024 |
| 中文关键词: 前列腺 骨盆 深度学习 分割 |
| 英文关键词:prostate pelvis deep learning segmentation |
| 基金项目:首都卫生发展科研专项（首发2020-2-40710）、北京大学第一医院科研种子基金（2020SF17）。 |
|
| 摘要点击次数: 1371 |
| 全文下载次数: 532 |
| 中文摘要: |
| 目的 评估3D U-Net深度学习（DL）模型基于盆腔T2WI自动分割盆腔软组织结构的可行性。方法 回顾性分析147例经病理证实或盆腔MRI随访观察确诊的前列腺癌或良性前列腺增生患者，其中28例接受2次、121例接受1次盆腔MR扫描，共175组T2WI；手动标注T2WI所示软组织结构，包括前列腺、膀胱、直肠、双侧精囊腺、尿道、双侧闭孔内肌及双侧耻骨直肠肌。按8：1：1比例将数据分为训练集（n=137）、调优集（n=21）和测试集（n=17），对3D U-Net分割模型进行训练。以手动标注结果为标准，根据测试集Dice相似系数（DSC）、Jaccard相似系数（JSC）、精确率（PRE）、召回率（REC）、准确率（ACC）及分割体积差异评价3D U-Net模型分割盆腔软组织结构的效能。结果 3D U-Net模型分割测试集盆腔各结构的DSC及JSC均>0.90，ACC、PRE和REC均>90.00%。3D U-Net模型分割的盆腔各结构体积与手动标注差异均无统计学意义（P均>0.05）。结论 3D U-Net DL模型可用于自动分割T2WI所示盆腔软组织结构。 |
| 英文摘要: |
| Objective To explore the feasibility of 3D U-Net deep learning (DL) model for automatic segmentation of pelvic soft tissue structures based on pelvic T2WI. Methods Pelvic MRI of 147 patients with pathologically or pelvic MRI follow-up confirmed prostate cancer or benign prostatic hyperplasia were analyzed retrospectively, including 28 patients underwent 2 times and 121 underwent one time MR scanning, and 175 groups of T2WI were obtained. The pelvic soft tissue structures, i.e. prostate, bladder, rectum, bilateral seminal vesicles, urethra, bilateral obturator muscles and bilateral puborectalis on T2WI were manually labeled. Then the data were divided into training set (n=137), validation set (n=21) and test set (n=17) at the ratio of 8:1:1, and the 3D U-Net segmentation model was trained. Taken manual annotation results as the reference standards, the segmentation performance of 3D U-Net DL model was evaluated according to differences of indexes including Dice similarity coefficient (DSC), Jaccard similarity coefficient (JSC), precision (PRE), recall (REC), accuracy (ACC) and segmentation volume in test set. Results DSC and JSC of 3D U-Net DL model segmentation of pelvic structures in test set were both >0.90,so were ACC, PRE and REC. The segmentation volumes using 3D U-Net DL model were not statistically different with those of manual annotation (all P>0.05). Conclusion 3D U-Net DL model could be used for automatic segmentation of pelvic soft tissue structures on T2WI. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
