Ecotourism requires the harmony of all factors involved in tourism system for a multilateral benefit. Based on such cognition, a concept of deep ecotourism development is put forward which includes two connotations: o...Ecotourism requires the harmony of all factors involved in tourism system for a multilateral benefit. Based on such cognition, a concept of deep ecotourism development is put forward which includes two connotations: on the one hand, it should give prominence to the display of the eco-culture of the tourist destination and tourists' eco-experience, in which way the development behavior on the tourist destination and the tourists' behavior will be regulated; on the other hand, it implies the deep harmony among tourist entrepreneurs and tourists, the local governments and the local residents, as well as tourist activities and the ecological environment in the tourism development for the multilateral benefit of every element involved and sustainable tourism development. The common sense is that the degrees in a certain tourism destination will differ and that consequently four levels of ecotourism are divided - very shallow ecotourism, shallow ecotourism, deep ecotourism and very deep ecotourism. To move shallow ecotourism toward deep one, two models of "four subjects and two wings" and "connecting the two wings" of deep ecotourism development system are introduced to make ecotourism industry favorable to the display of eco-culture and the sustainable development of the destination community. With the two models, a case study of ecotourism development in Louguantai National Forest Park was made as a demonstration. The ultimate purpose is to build an ideal new Shangri-La.展开更多
In recent years,Henan sticks to the points of General Secretary Xi Jinping,makes full use of its rich resources and relations,makes efforts to deepen friendship with people around the world,promotes relations with for...In recent years,Henan sticks to the points of General Secretary Xi Jinping,makes full use of its rich resources and relations,makes efforts to deepen friendship with people around the world,promotes relations with foreign countries,pushes展开更多
目的开发一种深度学习系统用于成人发育性髋关节发育不良(Developmental dysplasia of the hip,DDH)患者的Crowe分型辅助诊断,并且分析该系统对于帮助临床医学生掌握DDH分型的可行性。方法纳入149例X线片训练集、42例测试集以及21例验证...目的开发一种深度学习系统用于成人发育性髋关节发育不良(Developmental dysplasia of the hip,DDH)患者的Crowe分型辅助诊断,并且分析该系统对于帮助临床医学生掌握DDH分型的可行性。方法纳入149例X线片训练集、42例测试集以及21例验证集,分割盆骨、提取DDH局部图像块,将金标准结果与医学生、AI辅助医学生评估结果进行比较。结果测试集共纳入42例,其中女性30例,男性12例,年龄(69±12)岁,涉及发育不良髋关节67侧(左30侧,右37侧)。AI、医学生、AI辅助医学生评估结果与金标准的相关性为0.906[95%CI(0.850,0.941)]、0.823[95%CI(0.726,0.887)]、0.886[95%CI(0.821,0.929)];准确率分别为0.87、0.78、0.88;精确度分别为0.88、0.83、0.89;召回率分别为0.87、0.78、0.88;F1值分别为0.87、0.80、0.88。混淆矩阵和条件概率结果显示,预测准确率Ⅰ型DDH三组分别为0.98、0.88、0.96,Ⅱ型DDH三组分别为0.40、0.20、0.40,Ⅲ型DDH三组分别为0.56、0.67、0.78;Ⅳ型DDH三组分别为0.88、0.75、0.88。结论深度学习辅助诊断系统可以有效提高医学生对于DDH分型的评估能力,可作为医学生学习掌握DDH影像诊断的培训工具。展开更多
使用骨盆X光片诊断发育性髋关节发育不良(Developmental Dysplasia of the Hip,DDH)要求准确地标注髋关节关键点,而深度学习方法能作为可靠的辅助工具。针对骨盆片拍摄姿势和拍摄距离多样化问题,本文基于U-Net提出了RKD-UNet来检测髋关...使用骨盆X光片诊断发育性髋关节发育不良(Developmental Dysplasia of the Hip,DDH)要求准确地标注髋关节关键点,而深度学习方法能作为可靠的辅助工具。针对骨盆片拍摄姿势和拍摄距离多样化问题,本文基于U-Net提出了RKD-UNet来检测髋关节关键点。该模型使用残差块改进U-Net的卷积层和skip-connection路径,并将坐标注意力引入到编码器中以增强模型对关键点邻域的特征提取能力。在编码器顶部使用卷积和ASPP模块构成Bridge块,以[3,6,9]的空洞率融合不同尺度的特征信息并提升模型的感受野。本文使用包含骨盆正位片、蛙位片、下肢全长片和术后骨盆片的数据集训练和测试模型。RKD-UNet实现了3.19±2.19 px的平均关键点检测误差和2.83°±2.59°的平均髋臼角测量误差。对正常、轻度、中度和重度脱位案例诊断的F1分数分别达到89.6、77.1、57.9和94.1,高于医生的手动诊断结果。实验结果表明,RKD-UNet能准确检测髋关节关键点并辅助医生诊断DDH。展开更多
文摘Ecotourism requires the harmony of all factors involved in tourism system for a multilateral benefit. Based on such cognition, a concept of deep ecotourism development is put forward which includes two connotations: on the one hand, it should give prominence to the display of the eco-culture of the tourist destination and tourists' eco-experience, in which way the development behavior on the tourist destination and the tourists' behavior will be regulated; on the other hand, it implies the deep harmony among tourist entrepreneurs and tourists, the local governments and the local residents, as well as tourist activities and the ecological environment in the tourism development for the multilateral benefit of every element involved and sustainable tourism development. The common sense is that the degrees in a certain tourism destination will differ and that consequently four levels of ecotourism are divided - very shallow ecotourism, shallow ecotourism, deep ecotourism and very deep ecotourism. To move shallow ecotourism toward deep one, two models of "four subjects and two wings" and "connecting the two wings" of deep ecotourism development system are introduced to make ecotourism industry favorable to the display of eco-culture and the sustainable development of the destination community. With the two models, a case study of ecotourism development in Louguantai National Forest Park was made as a demonstration. The ultimate purpose is to build an ideal new Shangri-La.
文摘In recent years,Henan sticks to the points of General Secretary Xi Jinping,makes full use of its rich resources and relations,makes efforts to deepen friendship with people around the world,promotes relations with foreign countries,pushes
文摘目的开发一种深度学习系统用于成人发育性髋关节发育不良(Developmental dysplasia of the hip,DDH)患者的Crowe分型辅助诊断,并且分析该系统对于帮助临床医学生掌握DDH分型的可行性。方法纳入149例X线片训练集、42例测试集以及21例验证集,分割盆骨、提取DDH局部图像块,将金标准结果与医学生、AI辅助医学生评估结果进行比较。结果测试集共纳入42例,其中女性30例,男性12例,年龄(69±12)岁,涉及发育不良髋关节67侧(左30侧,右37侧)。AI、医学生、AI辅助医学生评估结果与金标准的相关性为0.906[95%CI(0.850,0.941)]、0.823[95%CI(0.726,0.887)]、0.886[95%CI(0.821,0.929)];准确率分别为0.87、0.78、0.88;精确度分别为0.88、0.83、0.89;召回率分别为0.87、0.78、0.88;F1值分别为0.87、0.80、0.88。混淆矩阵和条件概率结果显示,预测准确率Ⅰ型DDH三组分别为0.98、0.88、0.96,Ⅱ型DDH三组分别为0.40、0.20、0.40,Ⅲ型DDH三组分别为0.56、0.67、0.78;Ⅳ型DDH三组分别为0.88、0.75、0.88。结论深度学习辅助诊断系统可以有效提高医学生对于DDH分型的评估能力,可作为医学生学习掌握DDH影像诊断的培训工具。
文摘使用骨盆X光片诊断发育性髋关节发育不良(Developmental Dysplasia of the Hip,DDH)要求准确地标注髋关节关键点,而深度学习方法能作为可靠的辅助工具。针对骨盆片拍摄姿势和拍摄距离多样化问题,本文基于U-Net提出了RKD-UNet来检测髋关节关键点。该模型使用残差块改进U-Net的卷积层和skip-connection路径,并将坐标注意力引入到编码器中以增强模型对关键点邻域的特征提取能力。在编码器顶部使用卷积和ASPP模块构成Bridge块,以[3,6,9]的空洞率融合不同尺度的特征信息并提升模型的感受野。本文使用包含骨盆正位片、蛙位片、下肢全长片和术后骨盆片的数据集训练和测试模型。RKD-UNet实现了3.19±2.19 px的平均关键点检测误差和2.83°±2.59°的平均髋臼角测量误差。对正常、轻度、中度和重度脱位案例诊断的F1分数分别达到89.6、77.1、57.9和94.1,高于医生的手动诊断结果。实验结果表明,RKD-UNet能准确检测髋关节关键点并辅助医生诊断DDH。
