文章摘要
刘梦瑶,刘茹涵,姚一静,余倩,高乙惠,王芮,盛斌,姜立新.应用DeeplabV3+网络实现小儿髋关节超声图像识别[J].声学技术,2022,41(2):235~239
应用DeeplabV3+网络实现小儿髋关节超声图像识别
Application of DeeplabV3+ network in ultrasonic image recognition of pediatric hip joint
投稿时间:2021-10-08  修订日期:2021-11-03
DOI:10.16300/j.cnki.1000-3630.2022.02.013
中文关键词: 发育性髋关节发育不良  超声  图像分割  网络模型  DeeplabV3+
英文关键词: developmental dysplasia of the hip (DDH)  ultrasound  segmentation  network model  DeeplabV3+
基金项目:国家自然科学基金(81771850)资助项目。
作者单位E-mail
刘梦瑶 上海交通大学医学院附属仁济医院超声医学科, 上海 200127  
刘茹涵 上海交通大学电子信息与电气工程学院计算机系, 上海 200240  
姚一静 上海交通大学医学院附属仁济医院超声医学科, 上海 200127  
余倩 上海交通大学附属第六人民医院, 上海 200233  
高乙惠 上海交通大学附属第六人民医院, 上海 200233  
王芮 上海交通大学医学院附属仁济医院超声医学科, 上海 200127  
盛斌 上海交通大学电子信息与电气工程学院计算机系, 上海 200240  
姜立新 上海交通大学医学院附属仁济医院超声医学科, 上海 200127 jinger_28@sina.com 
摘要点击次数: 54
全文下载次数: 42
中文摘要:
      利用Graf法进行发育性髋关节发育不良(Developmental Dysplasia of the Hip,DDH)诊断时主要依靠骨-软骨交界面、股骨头、滑膜皱襞、关节囊及软骨膜、盂唇、软骨顶、骨性顶这7个解剖结构进行解剖验证,而初级医生对上述结构识别困难,因此文章提出了一种基于DeeplabV3+的网络模型用于7个结构的分割识别。首先对纳入的106例图像进行手动标记和预处理,之后将其分别输入DeeplabV3+和U-Net两种网络模型中,最终对其预测图表现和分割性能进行比较。与目前DDH图像分割中常用且表现优越的U-Net网络相比,DeeplabV3+网络的预测图包含的结构较多,边界分割也较清晰,其图像分割评价指标如相似性系数、豪斯多夫距离和平均豪斯多夫距离平均值的表现也优于U-Net网络。文章利用DeeplabV3+网络实现了DDH超声图像的7个结构分割,对临床医生进行后续图像的角度测量和分型诊断具有重要意义。
英文摘要:
      The diagnosis of developmental dysplasia of the hip (DDH) by Graf method mainly depends on the seven key structures, involving chondro-osseous border, femoral head, cartilagineous roof, synovial fold, labrum, bony roof and joint capsule. However, it is difficult for junior doctors to identify these structures. Therefore, a network model based on Deeplabv3+ for the segmentation of these seven structures is proposed in this paper. Firstly, 106 images were manually labeled and preprocessed, and then they were input into Deeplabv3+ and U-net network models respectively. Finally, the segmentation performances were compared. Compared with U-Net network which is commonly used and well-behaved in DDH segmentation, the image prediction of Deeplabv3+ network contain more structures and clearer boundary, and the main evaluation indices of segmentation, such as the average values of dice similarity coefficient, Hausdorff distance, average Hausdoff distance, also showed a better performance. The Deeplabv3+ network is first used to achieve segmentation of seven structures in DDH ultrasound images, which is of great significance for angle measurement and classification diagnosis.
查看全文   查看/发表评论  下载PDF阅读器
关闭