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...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。展开更多
The increasing prevalence of diabetes has become a global public health concern in the 21st century.In 2021,it was estimated that 537 million people had diabetes,and this number is projected to reach 643 million by 20...The increasing prevalence of diabetes has become a global public health concern in the 21st century.In 2021,it was estimated that 537 million people had diabetes,and this number is projected to reach 643 million by 2030,and 783 million by 2045[1].Such a huge burden of diabetes brings great challenges in its prevention and management,including early diagnosis,timely interventions,and regular monitoring of risk factor control and complications screening.Continuous self-care support and patient empowerment can enhance clinical and psychobehavioural outcomes[2],although these require additional resources including manpower,infrastructure(hard and technology),and finances.The emergence of digital health technologies(DHTs),especially artificial intelligence(AI),may help address these obstacles and alleviate the burden of diabetes[3].Large language models(LLMs),a generative AI that can accept image and text inputs and produce text outputs,have shown promise in various aspects of medical care.展开更多
文摘利用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。
基金supported by the National Key R&D Program of China(2022YFC2502800 and 2022YFC2407000)the National Natural Science Foundation of China(8238810007,82022012,81870598 and 62272298)+3 种基金the Shanghai Municipal Key Clinical SpecialtyShanghai Research Center for Endocrine and Metabolic Diseases(2022ZZ01002)the Chinese Academy of Engineering(2022-XY-08)the Innovative Research Team of High-level Local Universities in Shanghai(SHSMU-ZDCX20212700)。
文摘The increasing prevalence of diabetes has become a global public health concern in the 21st century.In 2021,it was estimated that 537 million people had diabetes,and this number is projected to reach 643 million by 2030,and 783 million by 2045[1].Such a huge burden of diabetes brings great challenges in its prevention and management,including early diagnosis,timely interventions,and regular monitoring of risk factor control and complications screening.Continuous self-care support and patient empowerment can enhance clinical and psychobehavioural outcomes[2],although these require additional resources including manpower,infrastructure(hard and technology),and finances.The emergence of digital health technologies(DHTs),especially artificial intelligence(AI),may help address these obstacles and alleviate the burden of diabetes[3].Large language models(LLMs),a generative AI that can accept image and text inputs and produce text outputs,have shown promise in various aspects of medical care.
