“傻瓜式”医学影像深度学习分析平台的设计与开发  

Design and Development of“Fool Type”Medical Imaging Deep Learning Analysis Platform

作　　者：贾楠 马宁[1] 吴燕 李燕[2] JIA Nan;MA Ning;WU Yan;LI Yan(Department of Computer Science and Technology,Baotou Medical College,Inner Mongolia University of Science and Technology,Baotou 014040,China;Clinical Laboratory of Baotou Central Hospital,Baotou 014040,China)

机构地区：[1]内蒙古科技大学包头医学院计算机科学与技术学院,内蒙古包头014040 [2]包头市中心医院检验科,内蒙古包头014040

出　　处：《现代信息科技》2023年第10期18-21,共4页Modern Information Technology

基　　金：内蒙古自治区高等学校科学研究项目(NJZY22050);包头医学院研究基金项目(BYJJ-KCRH202206)。

摘　　要：文章构建了一个针对医学影像的低代码深度学习实验平台。该平台可以处理2D、3D医学影像数据,实现基于深度学习的医学影像分类任务。该平台由数据集整理、训练集和测试集划分、模型训练、模型推理、结果展示5部分构成。每一部分均不需要编程,只需要指定数据路径或简单编写配置文件即可。文章还以一个新冠病毒感染智能检测的案例贯穿始终。该平台能够方便学生开展“智慧医疗”类创新创业项目,培养学生的创新能力和动手实践能力。This paper constructs a low code deep learning experimental platform for medical imaging.The platform can process 2D and 3D medical imaging data and realize medical imaging classification task based on deep learning.The platform consists offive parts:data set sorting,training set and test set division,model training,model reasoning,and result display.No programming is required for each part,just specify the data path or simply write the configurationfile is well.This paper also uses a case of intelligent detection of COVID-19 throughout.The platform can facilitate students to carry out“Wise Information Technology of Med”innovation and entrepreneurship projects,and cultivate students'innovation ability and practical ability.

关 键 词：智慧医疗 深度学习 医学影像 创新创业 实验平台 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论] TP18[自动化与计算机技术—计算机科学与技术] G482[文化科学—教育学]