面向轻量化医学图像分割网络的神经结构搜索  

Neural Architecture Search for Light-weight Medical Image Segmentation Network

作　　者：张福昌 仲国强 毛玉旭 ZHANG Fu-chang;ZHONG Guo-qiang;MAO Yu-xu(College of Information Science and Engineering,Ocean University of China,Qingdao,Shandong 266100,China)

机构地区：[1]中国海洋大学信息科学与工程学部,山东青岛266100

出　　处：《计算机科学》2022年第10期183-190,共8页Computer Science

基　　金：国家重点研发计划(2018AAA0100400);装备预研教育部联合基金(6141A020337);山东省自然科学基金(ZR2020MF131);福建省医疗数据挖掘与应用工程技术研究中心开放课题(MDM2018007);青岛市科技计划(21-1-4-ny-19-nsh)。

摘　　要：现有的性能优异的医学图像分割模型大都由领域专家手动设计,设计过程往往需要大量专业知识和反复实验。此外,过度复杂的分割模型不仅对硬件资源有较高要求,且分割效率较低。为此,提出了用于自动构建轻量化医学图像分割网络的神经结构搜索方法Auto-LW-MISN(Automatically Light-Weight Medical Image Segmentation Network)。通过构建轻量级搜索空间、设计适用于医学图像分割的搜索超网络、设计添加复杂性约束的可微分搜索策略,建立用于自动搜索轻量化医学图像分割网络的神经结构搜索框架。在显微镜细胞图像、肝脏CT图像和前列腺MR图像等数据集上进行实验,结果表明,Auto-LW-MISN能够针对不同模态的医学图像自动构建轻量化的分割模型,其分割精度相比U-net, Attention U-net, Unet++和NAS-Unet等方法均有提高。Most of the existing medical image segmentation models with excellent performance are manually designed by domain experts.The design process usually requires a lot of professional knowledge and repeated experiments.In addition, the over complex segmentation model not only has high requirements for hardware resources, but also has low segmentation efficiency.An neural architecture search method named Auto-LW-MISN(Automatically Light-weight Medical Image Segmentation Network) is proposed for automatic construction of light-weight medical image segmentation network.In this paper, by constructing a light-weight search space, designing a search super network for medical image segmentation, and designing a differentiable search stra-tegy with complexity constraints, a neural architecture search framework for automatic search of light-weight medical image segmentation network is established.Experimental results on microscope cell images, liver CT images and prostate MR images show that Auto-LW-MISN can automatically construct light-weight segmentation models for different modes of medical images, and its segmentation accuracy is improved compared with U-net, Attention U-net, Unet++and NAS-Unet.

关 键 词：深度学习 可微分神经结构搜索 轻量化卷积神经网络 自动化网络结构设计 医学图像分割 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]