基于可变形注意力transformer的胃肠癌病理图像细胞核自动分割方法  

Automatic nuclei segmentation of gastrointestinal cancer pathological images based on deformable attention transformer

作　　者：唐智贤 李镇 郭俏 胡家祺 王雪 姚旭峰 TANG Zhi-xian;LI Zhen;GUO Qiao;HU Jia-qi;WANG Xue;YAO Xu-feng(Department of Medical Imaging Technology,College of Medical Imaging,Shanghai University of Medicine and Health Sciences,Shanghai 201318,China;Institute of Biomedical Engineering,School of Health Science and Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China;Department of Pathology,Ruijin Hospital,Shanghai Jiao Tong University School of Medicine,Shanghai 200025,China)

机构地区：[1]上海健康医学院医学影像学院医学影像技术教研室,上海201318 [2]上海理工大学健康科学与工程学院生物医学工程研究所,上海200093 [3]上海交通大学医学院附属瑞金医院病理科,上海200025

出　　处：《复旦学报（医学版）》2024年第3期396-403,共8页Fudan University Journal of Medical Sciences

基　　金：上海市青年科技英才扬帆计划(21YF1418600);上海市高校青年教师培养资助计划(ZZ202216021);国家自然科学基金(61971275)。

摘　　要：目的使用深度学习算法实现胃肠癌病理图像的细胞核自动分割,辅助后续病理图像的定量分析。方法以2022年1月—2022年2月在上海交通大学医学院附属瑞金医院就诊的59例胃肠癌患者为研究对象,采用python和LabelMe对患者的病理图像进行数据脱敏、图片切割和感兴趣区域标注,共纳入944张病理图像,标注了9703个细胞核。通过构建一种基于深度学习的新型语义分割模型,模型引入可变型注意力transformer(deformable attention transformer,DAT),实现了病理图像细胞核自动、精准、高效分割,并采用多种分割评价标准评估模型性能。结果模型分割结果的平均绝对误差值(mean absolute error,MAE)为0.1126,骰子系数(dice coefficient,Dice)为0.7215,其效果明显优于U-net基线模型,并领先于ResU-net++、R2Unet和R2AttUnet等模型,且分割结果相对稳定,泛化性好。结论本研究建立的分割模型能够精准识别并分割出病理图像中的细胞核,鲁棒性和泛化性较好,有助于在实际应用中辅助诊断。Objective To achieve automatic segmentation of cell nuclei in gastrointestinal cancer pathological images by using a deep learning algorithm,so as to assist in the quantitative analysis of subsequent pathological images.Methods A total of 59 patients with gastrointestinal cancer treated in Ruijin Hospital,Shanghai Jiao Tong University School of Medicine from Jan 2022 to Feb 2022,were selected as the research objects.Python and LabelMe were used for data anonymization,image segmentation,and region of interest annotation of patients’pathological images.A total of 944 pathological images were included,and 9703 nuclei were annotated.Then,a new semantic segmentation model based on deep learning was constructed.The model introduced deformable attention transformer(DAT)to realize automatic,accurate and efficient segmentation of pathological image nuclei.Finally,multiple segmentation evaluation criteria are used to evaluate the model’s performance.Results The mean absolute error of the segmentation results of the model proposed in this paper was 0.1126,and the dice coefficient(Dice)was 0.7215.Its effect was significantly better than the U-net baseline model,and it was ahead of models such as ResU-net++,R2Unet and R2AttUnet.Moreover,the segmentation results were relatively stable with good generalization.Conclusion The segmentation model established in this study can accurately identify and segment the nuclei in the pathological images,with good robustness and generalization,which is helpful to play an auxiliary diagnostic role in practical applications.

关 键 词：深度学习模型 病理图像 细胞核分割 胃肠癌 诊断 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术] R-331[自动化与计算机技术—计算机科学与技术]