边缘计算在病理图像识别方向的应用  

Application of edge computing in pathological image recognition

作　　者：孙涛 王书浩 王伟 SUN Tao;WANG Shuhao;WANG Wei(School of Computer and Information Technology,Beijing Jiaotong University,Beijing 100080,China;Thorough Lab,Beijing Thorough Future Technology Co.,Ltd,Beijing 100080,China)

机构地区：[1]北京交通大学计算机与信息技术学院,北京100080 [2]北京透彻未来科技有限公司透彻实验室,北京100080

出　　处：《中国医学物理学杂志》2025年第3期328-335,共8页Chinese Journal of Medical Physics

基　　金：国家自然科学基金(61901436)。

摘　　要：病理图像是医学图像的重要组成部分,对于癌症等疾病的早期诊断至关重要。本文提出一种对医学病理图像进行识别分析的边缘推理解决方案,将边缘计算引入病理图像的研究方向,旨在利用边缘计算的优势,结合深度学习方法,实现实时的病理图像检测。实验结果表明,此方案的应用可以极大地保护患者隐私、提高病例图像处理的效率和实时性,并节省成本,具有很高的实际应用价值,可为医学诊断和治疗提供更快速和更准确的支持。Pathology images,being an essential part of medical imaging,are critical for early diagnosis of diseases such as cancer.The study introduces edge computing into pathology image research,and presents an edge inference solution for medical pathological image recognition and analysis,aiming to leverage the advantages of edge computing and combine with deep learning methods to achieve real-time pathology image detection.Experimental results show that the solution protects patient privacy well,improves the efficiency and real time of medical image processing,and greatly reduces costs,demonstrating that it has high practical application value and can provide more and efficient accurate support for medical diagnosis and treatment.

关 键 词：边缘计算 病理图像 深度学习 

分 类 号：R318[医药卫生—生物医学工程] TP3-05[医药卫生—基础医学]