基于深度学习的皮肤烧烫伤创面图像分割与分类及检测的研究进展  

Advances in deep learning-based image segmentation,classification and detection of skin burn and scald wounds

作　　者：王瑞[1] 张嘉炜 张克诚 周增丁[2] Rui Wang;Jiawei Zhang;Kecheng Zhang;Zengding Zhou(School of Communication and Information Engineering,Shanghai University,Shanghai 200444,China;Department of Burn and Plastic Surgery,Ruijin Hospital,School of Medicine,Shanghai Jiao Tong University,Shanghai 200025,China)

机构地区：[1]上海大学通信与信息工程学院,200444 [2]上海交通大学医学院附属瑞金医院烧伤整形科,200025

出　　处：《中华损伤与修复杂志（电子版）》2024年第2期172-175,共4页Chinese Journal of Injury Repair and Wound Healing(Electronic Edition)

基　　金：上海市科技创新行动计划自然科学基金(21ZR1440600)。

摘　　要：烧烫伤是常见的皮肤受损致伤原因,需高度重视。既往依靠烧伤专科医师临床经验判断,较难准确评估不同创面深浅等情况,由于光线、创面污秽等原因可能导致评估出现偏差。近年来,基于高级编程语言的深度学习在烧烫伤创面精准评估中的重要性逐步提高。本文通过回顾近几年不同研究人员将深度学习用于烧烫伤创面自动化诊断的研究进展,聚焦于创面图像分割、分类和检测三个关键方向,对不同深度学习技术在其中的应用进行总结和归纳,进一步探讨深度学习在患者烧烫伤创面自动化诊断方面面临的挑战,并对其未来应用前景进行展望。Burns and scalds are common causes of skin damage and injuries,and requiring significant attention.Traditionally,the assessment of burns relied on the clinical experience of burn specialists,making it challenging to accurately evaluate the depth and severity of different wounds.Factors like lighting and wound contamination could lead to assessment biases.In recent years,the importance of deep learning based on machine language in the precise evaluation of burn and scald wounds has gradually increased.This article reviews the progress made by scholars over the past few years in using deep learning for the automated diagnosis of burn and scald wounds,focusing on three key areas:wound image segmentation,classification,and detection.It summarizes and categorizes the application of different deep learning techniques in these areas,further explores the challenges faced in automated diagnosis of burn and scald wounds using deep learning,and looks forward to its future application prospects.

关 键 词：烧烫伤 深度学习 图像分割 图像分类与检测 

分 类 号：R644[医药卫生—外科学]