Gastric cancer pathological image diagnosis system based on ResNet and UNet

doi:10.3969/j.issn.1000-1565.2024.04.012

Abstract

Abstract: Considering that manual identification and diagnosis of gastric cancer pathological images may cause missed detection and in order to make diagnosis more accurate, a pathological image diagnosis system based on ResNet and UNet is proposed, aiming to classify, segment and output the diagnosis results of pathological images. The ResNet model is used to classify gastric cancer pathological images with and without cancer. The UNet model is improved, and the improved model adds a convolutional block attention module before each down-sampling and up-sampling to enhance the models attention to cancerous areas. The residual module is used to replace the two convolutions in the encoding part to improve feature utilization; and the Inception module is used to replace the two convolutions in the up-sampling- DOI:10.3969/j.issn.1000-1565.2024.04.012基于ResNet和UNet的胃癌病理图像诊断系统张文悦¹,贾子彦¹,李青²,张大川²,潘玲佼¹,沈大伟¹(1.江苏理工学院电气信息工程学院,江苏常州 213001;2.常州市第一人民医院病理科,江苏常州 213004)摘要:考虑到人工对胃癌病理图像的判别和诊断可能存在漏检的问题,为使诊断更加准确,提出一种基于ResNet和UNet的病理图像诊断系统,旨在实现对病理图像的分类、分割以及输出诊断结果.采用ResNet模型对胃癌病理图像进行有癌和无癌的分类.对UNet模型进行改进,改进后的模型在每个下采样和上采样之前加入卷积注意力模块,以增强模型对癌变区域的关注.使用残差模块替代编码部分的2次卷积,来提高特征的利用率;利用Inception模块来替代解码部分上采样中的2个卷积,从而扩充其宽度并获取不同尺度的特征.将分类与分割结果综合考虑,获取最终的胃癌病理图像的诊断结果.实验结果表明,该系统可以有效地诊断胃癌病理图像中是否存在癌变.关键词:病理图像;图像分类;UNet;图像分割;胃癌诊断中图分类号:TP391.4 文献标志码:A 文章编号:1000-1565(2024)04-0433-08Gastric cancer pathological image diagnosis system based on ResNet and UNetZHANG Wenyue¹, JIA Ziyan¹, LI Qing², ZHANG Dachuan², PAN Lingjiao¹, SHEN Dawei¹(1. School of Electrical and Information Engineering, Jiangsu University of Technology, Changzhou 213001, China; 2. Department of Pathology, Changzhou First Peoples Hospital, Changzhou 213004, China)Abstract: Considering that manual identification and diagnosis of gastric cancer pathological images may cause missed detection and in order to make diagnosis more accurate, a pathological image diagnosis system based on ResNet and UNet is proposed, aiming to classify, segment and output the diagnosis results of pathological images. The ResNet model is used to classify gastric cancer pathological images with and without cancer. The UNet model is improved, and the improved model adds a convolutional block attention module before each down-sampling and up-sampling to enhance the models attention to cancerous areas. The residual module is used to replace the two convolutions in the encoding part to improve feature utilization; and the Inception module is used to replace the two convolutions in the up-sampling- 收稿日期:2023-10-27;修回日期:2024-04-25 基金项目:国家自然科学基金资助项目(62001196); 江苏省“333高层次人才培养工程”项目(2022-3-4-107); 常州市科技计划项目(CM20223015); 常州应用基础研究项目(CJ20220064;CJ20220059) 第一作者:张文悦(1997—),女,江苏理工学院在读硕士研究生,主要从事计算机医学图像处理方向研究.E-mail:zhangwenyue97wren@163.com 通信作者:贾子彦(1981—),男,江苏理工学院副教授,主要从事可见光通信、机器视觉、5G方向研究.E-mail:jiaziyan@jsut.edu.cn第4期张文悦等:基于ResNet和UNet的胃癌病理图像诊断系统河北大学学报(自然科学版) 第44卷of the decoding part, thereby expanding its width to obtain features of different scales. The classification and segmentation results are comprehensively considered to obtain the final diagnostic results of gastric cancer pathological images. Experimental results show that this system can effectively diagnose the presence of cancer in gastric cancer pathological images.

Key words: pathological images, image classification, UNet, image segmentation, diagnosis of gastric cancer

CLC Number:

TP391.4

ZHANG Wenyue, JIA Ziyan, LI Qing, ZHANG Dachuan, PAN Lingjiao, SHEN Dawei. Gastric cancer pathological image diagnosis system based on ResNet and UNet[J]. Journal of Hebei University(Natural Science Edition), 2024, 44(4): 433-440.

References

[1] BRAY F, LAVERSANNE M, WEIDERPASS E, et al.The ever-increasing importance of cancer as a leading cause of premature death worldwide[J].Cancer, 2021, 127(16):3029-3030.DOI.org/10.1002/cncr.33587.
[2] 郑荣寿,张思维,孙可欣,等.2016年中国恶性肿瘤流行情况分析[J].中华肿瘤杂志, 2023, 45(3):212-220.DOI:10.3760/cma.j.cn112152-20220922-00647.
[3] 石田蕾,张家意,鲍泳扬,等.基于深度学习的黑色素细胞病变全流程智能化精准诊断[J].生物医学工程学杂志, 2022,39(05):919-927.DOI:10.7507/1001-5515.202203080.
[4] 陈颍锶,李晗,周雪婷,等.融合空洞卷积与注意力的胃癌组织切片分割[J].中国图象图形学报, 2021,26(09):2281-2292.DOI:10.11834/jig.200765.
[5] 王继伟,樊伟,陈岗,等.基于深度卷积神经网络的数字病理辅助诊断系统设计[J].中国数字医学, 2020,15(12)48-52.DOI:10.3969/j.issn.1673-7571.2020.12.014.
[6] ZHAO Y, ZHANG J, HU D, et al.Application of deep learning in histopathology images of breast cancer: a review[J]. Micromachines, 2022,13(12):2197. DOI: 10.3390/mi13122197.
[7] 喻殿智,张欣,迟杏.基于CA-DenseNet的乳腺癌病理图像识别[J].国外电子测量技术,2022,41(05):137-143. DOI:10.19652/j.cnki.femt.2103531.
[8] 杨昆,常世龙,王尉丞,等.基于sECANet通道注意力机制的肾透明细胞癌病理图像ISUP分级预测[J].电子与信息学报,2022,44(01):138-148. DOI: 10.11999/JEIT210900.
[9] LI X, WANG Y, TANG Q, et al. Dual U-Net for the segmentation of overlapping glioma nuclei[J]. IEEE Access, Vol84040-84052,2019. DOI: 10.1109/ACCESS.2019.2924744.
[10] BAYKAL E, DOAN H, EKINCI M, et al. An ensemble of fine-tuned fully convolutional neural networks for pleural effusion cell nuclei segmentation[J]. Computers and Electrical Engineering, 2019: 81. DOI: 10.1016/j.compeleceng.2019.106533.
[11] 邓健志,支佩佩,张峰铭,等.结合拆分注意力特征融合的病理图像分割网络[J].科学技术与工程,2023,23(7):2922-2931. DOI: 10.3969/j. issn.1671-1815.2023.07.030.
[12] ZHANG Y Y, LI Y Q, HE D Z. U-shaped network with multi-scale feature extraction for lesion region segmentation in gastric cancer images[C] //2023 IEEE 3rd International Conference on Information Technology, Big Data and Artificial Intelligence(ICIBA), Chongqing, China, 2023:234-238. DOI: 10.1109/ICIBA56860.2023.10165314.
[13] WARIN K, LIMPRASERT W, SUEBNUKARN S, et al. Performance of deep convolutional neural network for classification and detection of oral potentially malignant disorders in photographic images[J]. International journal of oral and maxillofacial surgery, 2022, 51(5):699-704. DOI:10.1016/j.ijom.2021.09.001.
[14] AZOUR F, BOUKERCHE A. An efficient transfer and ensemble learning based computer aided breast abnormality diagnosis system[J]. IEEE Access, vol. 11, 2023: 21199-21209. DOI: 10.1109/ACCESS.2022.3192857.
[15] HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition [J]. 2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR), Las Vegas, NV, USA, IEEE, 2016, 770-778. DOI:10.1109/CVPR.2016.90.
[16] RONNEBERGER O, FISCHER P, BROX T. U-Net: convolutional networks for biomedical image segmentation[J]. arXiv preprint arXiv:1505.04597,2015. DOI.org/10.48550/arXiv.1505.04597.
[17] WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module [C] // Proceedings of the European conference on computer vision(ECCV). 2018: 3-19. DOI: 10.1007/978-3-030-01234-2_1.
[18] CHEN L C, PAPANDREOU G, SCHROFF F, et al. Rethinking atrous convolution for semantic image segmentation[J]. arXiv preprint arXiv:1706.05587, 2017. DOI: 10.48550/ arXiv:1706.05587.
[19] ZHAO H, SHI J, QI X, et al. Pyramid scene parsing network [C] //Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 2881-2890. DOI: 10.1109/CVPR.2017.660.
[20] CHEN J, LU Y, YU Q, et al. TransUNet: transformers make strong encoders for medical image segmentation[J]. arXiv preprint arXiv:2102.04306, 2021. DOI: 10.48550/arXiv.2102.04306. (