腹腔镜手术自动化是智能外科的重要组成部分,其前提是腔镜视野下手术器械与脏器实时精准分割。受术中血液污染、烟雾干扰等复杂因素影响,器械与脏器实时精准分割面临巨大挑战,现有图像分割方法均表现不佳。因此提出一种基于注意力感知...腹腔镜手术自动化是智能外科的重要组成部分,其前提是腔镜视野下手术器械与脏器实时精准分割。受术中血液污染、烟雾干扰等复杂因素影响,器械与脏器实时精准分割面临巨大挑战,现有图像分割方法均表现不佳。因此提出一种基于注意力感知与空间通道的快速分割网络(ASC-Net),以实现腹腔镜图像中器械和脏器快速精准分割。在UNet架构下,设计了注意力感知与空间通道模块,通过跳跃连接将二者嵌入编码与解码模块间,使网络重点关注图像中相似目标间深层语义信息差异,同时多维度学习各目标的多尺度特征。此外,采用预训练微调策略,减小网络计算量。实验结果表明:在EndoVis2018数据集上的平均骰子系数(mDice)、平均重叠度(mIoU)、平均推理时间(mIT)分别为90.64%,86.40%和16.73 ms (60帧/秒),相比于现有最先进方法,mDice与mIoU提升了26%与39%,且mIT降低了56%;在AutoLaparo数据集上的mDice,mIoU和mIT分别为93.72%,89.43%和16.41ms(61帧/秒),同样优于对比方法。该方法在保证分割速度的同时有效提升了分割精度,实现了腹腔镜图像中手术器械和脏器的快速精准分割,将助力腹腔镜手术自动化快速发展。展开更多
本文设计了由四聚长方体组成的全介质超表面,其中每个长方体刻蚀两个椭圆柱并填装空气.当分别为超表面单独引入面内对称破缺、位移扰动和周期扰动时,可在近红外波段产生稳健的准连续域束缚态模式(quasi-bound states in the continuum)...本文设计了由四聚长方体组成的全介质超表面,其中每个长方体刻蚀两个椭圆柱并填装空气.当分别为超表面单独引入面内对称破缺、位移扰动和周期扰动时,可在近红外波段产生稳健的准连续域束缚态模式(quasi-bound states in the continuum).通过测量准BIC (quasi-BIC)模式的谐振波长,计算准BIC模式的Q因子(quality fector)与不对称参数的关系,可进一步证实不对称参数对准BIC共振频率和Q因子的可调谐性.在此基础上,当同时引入面内对称破缺、位移扰动和周期扰动时,可获得5个高Q因子的准BIC模式.共振峰的数量、位置以及Q因子都可通过调整面内破缺、位移扰动和周期扰动的程度进行调控.该超表面的设计可为传感器的多参数传感以及灵敏度等性能的提升提供一种全新思路.展开更多
Within the dicynodont genus Turfanodon,there are two recognized species,T.bogdaensis and T.jiufengensis.Both species are known by relatively complete cranial materials,but the mandibles and most postcranial bones have...Within the dicynodont genus Turfanodon,there are two recognized species,T.bogdaensis and T.jiufengensis.Both species are known by relatively complete cranial materials,but the mandibles and most postcranial bones have been described only for T.jiufengensis.This paper reports new dicynodont specimens from Turpan,Xinjiang,referring them to T.bogdaensis.They can clearly be differentiated from T.jiufengensis by the flatter lateral surface of the snout region,a prominent swelling on the lateral dentary shelf,and the rounded femoral head.The diagnosis of Turfanodon is revised.The combination of a flat circumorbital rim,posterior portion of anterior pterygoid rami with converging ventral ridges,and a possible autapomorphy,a deep notch on scapula forming procoracoid foramen,are confirmed.It also differentiated from all dicynodonts other than Myosaurus,Kembawacela and Lystrosaurus by having accessory ridges lateral to the median palatal ridge.展开更多
文摘腹腔镜手术自动化是智能外科的重要组成部分,其前提是腔镜视野下手术器械与脏器实时精准分割。受术中血液污染、烟雾干扰等复杂因素影响,器械与脏器实时精准分割面临巨大挑战,现有图像分割方法均表现不佳。因此提出一种基于注意力感知与空间通道的快速分割网络(ASC-Net),以实现腹腔镜图像中器械和脏器快速精准分割。在UNet架构下,设计了注意力感知与空间通道模块,通过跳跃连接将二者嵌入编码与解码模块间,使网络重点关注图像中相似目标间深层语义信息差异,同时多维度学习各目标的多尺度特征。此外,采用预训练微调策略,减小网络计算量。实验结果表明:在EndoVis2018数据集上的平均骰子系数(mDice)、平均重叠度(mIoU)、平均推理时间(mIT)分别为90.64%,86.40%和16.73 ms (60帧/秒),相比于现有最先进方法,mDice与mIoU提升了26%与39%,且mIT降低了56%;在AutoLaparo数据集上的mDice,mIoU和mIT分别为93.72%,89.43%和16.41ms(61帧/秒),同样优于对比方法。该方法在保证分割速度的同时有效提升了分割精度,实现了腹腔镜图像中手术器械和脏器的快速精准分割,将助力腹腔镜手术自动化快速发展。
文摘本文设计了由四聚长方体组成的全介质超表面,其中每个长方体刻蚀两个椭圆柱并填装空气.当分别为超表面单独引入面内对称破缺、位移扰动和周期扰动时,可在近红外波段产生稳健的准连续域束缚态模式(quasi-bound states in the continuum).通过测量准BIC (quasi-BIC)模式的谐振波长,计算准BIC模式的Q因子(quality fector)与不对称参数的关系,可进一步证实不对称参数对准BIC共振频率和Q因子的可调谐性.在此基础上,当同时引入面内对称破缺、位移扰动和周期扰动时,可获得5个高Q因子的准BIC模式.共振峰的数量、位置以及Q因子都可通过调整面内破缺、位移扰动和周期扰动的程度进行调控.该超表面的设计可为传感器的多参数传感以及灵敏度等性能的提升提供一种全新思路.
文摘Within the dicynodont genus Turfanodon,there are two recognized species,T.bogdaensis and T.jiufengensis.Both species are known by relatively complete cranial materials,but the mandibles and most postcranial bones have been described only for T.jiufengensis.This paper reports new dicynodont specimens from Turpan,Xinjiang,referring them to T.bogdaensis.They can clearly be differentiated from T.jiufengensis by the flatter lateral surface of the snout region,a prominent swelling on the lateral dentary shelf,and the rounded femoral head.The diagnosis of Turfanodon is revised.The combination of a flat circumorbital rim,posterior portion of anterior pterygoid rami with converging ventral ridges,and a possible autapomorphy,a deep notch on scapula forming procoracoid foramen,are confirmed.It also differentiated from all dicynodonts other than Myosaurus,Kembawacela and Lystrosaurus by having accessory ridges lateral to the median palatal ridge.
