改进卷积神经网络的前列腺癌格里森分级研究

doi:10.3778/j.issn.1002-8331.2009-0506

摘要/Abstract

摘要： 前列腺癌是全球范围内男性最常见的癌症之一，仅次于肺癌。在前列腺癌的诊断过程中最常用的方法是病理学专家通过显微镜对染色活检组织进行观察，得出组织微阵列图像的Gleason评分。在大量的组织微阵列图像下，病理学专家使用Gleason模式对前列腺癌组织微阵列进行评分非常耗时，易受到不同观察者之间主观因素的影响，且可重复性低。深度学习和计算机视觉的发展使得病理学计算机辅助诊断系统更具有客观性和可重复性。U-Net是医学影像分割领域应用最广泛的的网络，不同于以往研究中使用分类器，提出了一种基于改进的U-Net网络的区域分割模型，通过密集连接块来融合深层和浅层特征的同时对各个尺度的特征进行监督。可以减少网络参数，提高计算效率，并在标注完整的数据集上验证了方法有效性。

关键词: U-Net网络, Gleason模式, 图像分割

Abstract: Prostate cancer is one of the most common cancers in the world, second only to lung cancer. In the diagnosis of prostate cancer, the most commonly used method is pathological experts to observe the stained biopsy tissue through the microscope, and get the Gleason score of tissue microarray image. In a large number of tissue microarray images, it is very time-consuming for pathological experts to use Gleason pattern to score prostate cancer tissue microarray, which is easily affected by subjective factors among different observers, and has low repeatability. The development of deep learning and computer vision makes the computer-aided diagnosis system of pathology more objective and repeatable. U-Net is the most widely used network in the field of medical image segmentation, which is different from the classifier used in previous studies. A region segmentation model based on improved net network is proposed, which combines the deep and shallow features through dense connection blocks and supervises the features of each scale at the same time. It can reduce the network parameters and improve the calculation efficiency, and verify the effectiveness of the method on the annotated complete dataset.

Key words: U-Net network, Gleason pattern, image segmentation

单怡晴, 黄梦醒, 张雨, 李玉春, 张新华, 冯思玲, 陈晶. 改进卷积神经网络的前列腺癌格里森分级研究[J]. 计算机工程与应用, 2022, 58(7): 243-249.

SHAN Yiqing, HUANG Mengxing, ZHANG Yu, LI Yuchun, ZHANG Xinhua, FENG Siling, CHEN Jing. Gleason Grading of Prostate Cancer Based on Improved Convolution Neural Network[J]. Computer Engineering and Applications, 2022, 58(7): 243-249.

参考文献

[1] 王宁，刘硕，杨雷，等.2018全球癌症统计报告解读[J].肿瘤综合治疗电子杂志，2019，5（1）：87-97.
WANG N，LIU S，YANG L，et al.Interpretation of 2018 global cancer statistical report[J].Journal of Comprehensive Cancer Treatment，2019，5（1）：87-97.
[2] JONATHAN I EPSTEIN，ALLSBROOK W C，et al.The 2005 international society of urological pathology（ISUP） consensus conference on gleason grading of prostatic carcinoma[J].The American Journal of Surgical Pathology，2005，29（9）：1228-1242.
[3] GLEASON D F，GLEASON D F.Classification of prostatic carcinomas：Cancer chemother rep 50：125-128[R].
[4] 梁妙兴，陈志坚，王宇，等.前列腺癌穿刺引导器穿刺活检标本与前列腺全切标本病理Gleason分级相关性研究[J].医学食疗与健康，2020，18（13）：1-3.
LIANG M X，CHEN Z J，WANG Y，et al.Study on the correlation between pathological Gleason classification of prostate cancer biopsy specimens and total prostatectomy specimens[J].Medical Dietotherapy and Health，2020，18（13）：1-3.
[5] CHAURASIA A，CULURCIELLO E.LinkNet：Exploiting encoder representations for efficient semantic segmentation[C]//2017 IEEE Conference on Visual Communications and Image Processing，2017：1-4.
[6] KOUROSH J K，HAMID S Z.Multiwavelet grading of pathological images of prostate[J].IEEE Transactions on Bio-medical Engineering，2003，50（6）：840-851.
[7] SMITH Y，ZAJICEK G，WERMAN M，et al.Similarity measurement method for the classification of architecturally differentiated images[J].Journal of Biomedical Informatics，1999，32（1）：1-12.
[8] 段峰，王耀南，雷晓峰，等.机器视觉技术及其应用综述[J].自动化博览，2002（3）：62-64.
DUAN F，WANG Y N，LEI X F，et al.Overview of machine vision technology and its application[J].Automation Expo，2002（3）：62-64.
[9] 廖欣，郑欣，邹娟，等.基于深度卷积神经网络的宫颈细胞病理智能辅助诊断方法[J].液晶与显示，2018，33（6）：528-537.
LIAO X，ZHENG X，ZOU J，et al.Intelligent diagnosis method of cervical cell pathology based on deep convolution neural network[J].Liquid Crystal and Display，2018，33（6）：528-537.
[10] 郑光远，刘峡壁，韩光辉.医学影像计算机辅助检测与诊断系统综述[J].软件学报，2018，29（5）：1471-1514.
ZHENG G Y，LIU X B，HAN G H.Review of computer aided detection and diagnosis system for medical imaging[J].Journal of Software，2018，29（5）：1471-1514.
[11] ESTEVA A，KUPREL B，NOVOA R A，et al.Dermatologist-level classification of skin cancer with deep neural networks[J].Nature，2017，542：115-118.
[12] DE FAUW J，LEDSAM J R，ROMERA-PAREDES B，et al.Clinically applicable deep learning for diagnosis and referral in retinal disease[J].Nature Medicine，2018，24（9）：1342-1350.
[13] SHELHAMER E，LONG J，DARRELL T.Fully convolutional networks for semantic segmentation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2017，39（4）：640-651.
[14] RONNEBERGER O，FISCHER P，BROX T.U-Net：Convolutional networks for biomedica image segmentation[C]//Proceedings of Internatinal Conference on Medical Image Computing and Computer-Assisted Intervetion，2015：234-241.
[15] 肖朝霞，陈胜.图像语义分割问题研究综述[J].软件导刊，2018，17（8）：6-8.
XIAO C X，CHEN S.A survey of image semantic segmentation[J].Software Guide，2018，17（8）：6-8.
[16] 王孙平，陈世峰.融合深度图像的卷积神经网络语义分割方法[J].集成技术，2018，7（5）：58-66.
WANG S P，CHEN S F.Convolutional neural network semantic segmentation method based on fusion of depth image[J].Ensemble Technology，2018，7（5）：58-66.
[17] IRIGARAY P.Re：Prostate cancer diagnosis and treatment after the introduction of prostate-specific antigen screening：1986—2005[J].Journal of the National Cancer Institute，2010，102（7）：506-507.
[18] WILSON M L，FLEMING K A，KUTI M A，et al.Access to pathology and laboratory medicine services：A crucial gap[J].The Lancet，2018，391：1927-1938.
[19] 韩超，朱丽娜，刘想，等.基于U-Net实现前列腺MR图像上腺体的自动分割和径线测量：临床植入验证研究[J].放射学实践，2020，35（4）：519-524.
HAN C，ZHU L N，LIU X，et al.Automatic segmentation and diameter measurement of prostate gland on MR images based on U-Net：Clinical implant verification study[J].Radiology Practice，2020，35（4）：519-524.
[20] 颜野，夏海缀，李旭升，等.基于U型卷积神经网络学习的前列腺癌影像重建模型在手术导航中的应用[J].北京大学学报（医学版），2019，51（3）：596-601.
YAN Y，XIA H Z，LI X S，et al.Application of prostate cancer image reconstruction model based on U-type convolutional neural network learning in surgical navigation[J].Journal of Peking University（Medical Edition），2019，51（3）：596-601.
[21] OZKAN T A，ERUYAR A T，CEBECI O O，et al.Interobserver variability in Gleason histological grading of prostate cancer[J].Scandinavian Journal of Urology & Nephrology，2016，50（6）：420-424.
[22] MILLETARI F，NAVAB N，AHMADI S.V-Net：Fully convolutional neural networks for volumetric medical image segmentation[C]//Proceedings of the 4th International Conference on 3D Vision，2016：565-569.