ACT-Net: Anchor-context Action Detection in Surgery Videos
University of Birmingham(伯明翰大学);Southern University of Science and Technology(南方科技大学);Research Institute of Trustworthy Autonomous Systems(可信自主系统研究院);National University of Singapore(新加坡国立大学);Third medical center of Chinese PLAGH(中国人民解放军总医院第三医学中心);Institute of High Performance Computing (IHPC), Agency for Science, Technology and Research (A*STAR)(新加坡科技研究局高性能计算研究所)
Luoying Hao, Yan Hu, Wenjun Lin, Qun Wang, Heng Li, Huazhu Fu, Jinming Duan, Jiang Liu
Recognition and localization of surgical detailed actions is an essential component of developing a context-aware decision support system. However, most existing detection algorithms fail to provide high-accuracy action classes even having their locations, as they do not consider the surgery procedure's regularity in the whole video. This limitation hinders their application. Moreover, implementing the predictions in clinical applications seriously needs to convey model confidence to earn entrustment, which is unexplored in surgical action prediction. In this paper, to accurately detect fine-grained actions that happen at every moment, we propose an anchor-context action detection network (ACTNet), including an anchor-context detection (ACD) module and a class conditional diffusion (CCD) module, to answer the following questions: 1) where the actions happen; 2) what actions are; 3) how confidence predictions are. Specifically, the proposed ACD module spatially and temporally highlights the regions interacting with the extracted anchor in surgery video, which outputs action location and its class distribution based on anchor-context interactions. Considering the full distribution of action classes in videos, the CCD module adopts a denoising diffusion-based generative model conditioned on our ACD estimator to further reconstruct accurately the action predictions. Moreover, we utilize the stochastic nature of the diffusion model outputs to access model confidence for each prediction. Our method reports the state-of-the-art performance, with improvements of 4.0% mAP against baseline on the surgical video dataset.
Attachment Anchors: A Novel Framework for Laparoscopic Grasping Point Prediction in Colorectal Surgery
Technical University of Munich, TUM School of Medicine and Health, TUM University Hospital rechts der Isar, Department of Surgery, Research Group MITI(慕尼黑技术大学,TUM医学与健康学院,TUM慕尼黑大学医院rechts der Isar,外科部,MITI研究组);Technical University of Munich, TUM School of Medicine and Health, TUM University Hospital rechts der Isar, Chair for AI in Healthcare and Medicine Munich(慕尼黑技术大学,TUM医学与健康学院,TUM慕尼黑大学医院rechts der Isar,人工智能在医疗与健康中的chair);Department of Computing, Imperial College London(伦敦帝国学院计算系)
SCPM-Net: An Anchor-free 3D Lung Nodule Detection Network using Sphere Representation and Center Points Matching
University of Electronic Science and Technology of China(电子科技大学);SenseTime Research(商汤科技研究院);Tongji University(同济大学);Sichuan University West China Hospital(四川大学华西医院);Rutgers University(罗格斯大学)
Detection and Localization of Robotic Tools in Robot-Assisted Surgery Videos Using Deep Neural Networks for Region Proposal and Detection
Georgia Institute of Technology(佐治亚理工学院);Twentieth Century Fox(二十世纪福克斯公司);Starfleet Academy(星际舰队学院);Tyrell Inc.(泰瑞尔公司)
机器人辅助手术(RAS)视频的视频理解是一个活跃的研究领域。对外科医生的手势和技能水平进行建模提出了一个有趣的问题。所得出的见解可以应用于有效的技能获取、客观的技能评估、实时反馈和人机协作手术。我们提出了一个解决方案的工具检测和本地化开放的问题,在RAS视频理解,使用严格的计算机视觉的方法和深度学习的最新进展。我们提出了一个架构,使用多模态卷积神经网络的快速检测和RAS视频中的工具定位。据我们所知,这种方法将是第一个将深度神经网络用于RAS视频中的工具检测和定位的方法。我们的架构应用区域建议网络(RPN),和多模态双流卷积网络的目标检测,共同预测对象和定位的融合图像和时间的运动线索。我们的平均精度(AP)为91%,每个测试帧检测的平均计算时间为0.1秒,这表明我们的研究优于传统的医学成像方法,同时也强调了使用RPN的精度和效率的好处。我们还介绍了一个新的数据集,ATLAS Dione,RAS视频理解。我们的数据集提供了来自Roswell Park Cancer Institute(RPCI)(Buffalo,NY)的十名外科医生在达芬奇手术系统(dVSS R)上执行六种不同手术任务的视频数据,每帧都有机器人工具的注释。
Effective semantic segmentation in Cataract Surgery: What matters most?
University College London (UCL)(伦敦大学学院);King’s College London (KCL)(伦敦国王学院);Moorfields Eye Hospital(摩尔菲尔兹眼科医院);Institute of Ophthalmology, University College London(伦敦大学学院眼科研究所);Wellcome/EPSRC Centre for Interventional and Surgical Sciences(惠康/英国工程与物理科学研究委员会介入与外科科学中心);School of Biomedical Engineering & Imaging Sciences(生物医学工程与影像科学学院)
RSDNet: Learning to Predict Remaining Surgery Duration from Laparoscopic Videos Without Manual Annotations
ICube(ICube实验室);University of Strasbourg(斯特拉斯堡大学);CNRS(法国国家科学研究中心);IHU Strasbourg(斯特拉斯堡大学医院研究所);University Hospital of Strasbourg(斯特拉斯堡大学医院);IRCAD(法国消化外科研究所)
Video-Instrument Synergistic Network for Referring Video Instrument Segmentation in Robotic Surgery
Hong Kong University of Science and Technology (Guangzhou)(香港科技大学(广州));Imperial College London(帝国理工学院);Hong Kong University of Science and Technology(香港科技大学);The General Hospital of Southern Theatre Command of PLA(中国人民解放军南部战区总医院);National University of Singapore(新加坡国立大学)
DALNet: A Rail Detection Network Based on Dynamic Anchor Line
School of Control Science and Engineering, Dalian University of Technology(大连理工大学控制科学与工程学院);Dalian Rail Transmit Intelligent Control and Intelligent Operation Technology Innovation Center(大连轨道交通智能控制与智能运维技术创新中心);Dalian Seasky Automation Co., Ltd(大连海空自动化有限公司)
Future-State Predicting LSTM for Early Surgery Type Recognition
ICube, University of Strasbourg, CNRS, IHU Strasbourg(斯特拉斯堡大学ICube、法国国家科学研究中心、斯特拉斯堡大学医院研究所);University Hospital of Strasbourg(斯特拉斯堡大学医院);IRCAD(消化癌症外科研究所);IHU Strasbourg(斯特拉斯堡大学医院研究所)
Anchor Retouching via Model Interaction for Robust Object Detection in Aerial Images
Nanjing University of Aeronautics and Astronautics(南京航空航天大学);National Research Lobachevsky State University of Nizhny Novgorod(下诺夫哥罗德国立研究型罗巴切夫斯基大学);University of Leicester(莱斯特大学)
目标检测在计算机视觉中取得了巨大的进步。外观退化的小目标检测是一个突出的挑战,特别是对于航空观测。为了收集足够的正/负样本进行启发式训练,大多数对象检测器都会预设区域锚点,以便根据地面真实数据计算Intersection-over-Union(IoU)。在这种情况下,小物体经常被遗弃或贴错标签。本文提出了一种有效的动态增强锚(DEA)网络来构造一种新的训练样本生成器。与其他最先进的技术不同,所提出的网络利用一个样本筛选器来实现基于锚的单元和无锚单元之间的交互式样本筛选,以生成合格的样本。此外,多任务联合训练与保守的基于锚点的推理方案,提高了所提出的模型的性能,同时降低了计算复杂性。该方案支持定向和水平目标检测任务。在两个具有挑战性的空中基准上进行了广泛的实验(即,DOTA和HRSC 2016)表明,我们的方法在准确性方面达到了最先进的性能,具有中等的推理速度和训练的计算开销。在DOTA上,我们的DEA-Net集成了RoI-Transformer的基线,对于具有较弱骨干网络的定向目标检测(ResNet-101 vs ResNet-152),我们的DEA-Net超过了0.40%的平均平均精度(mAP),对于具有相同骨干网络的水平目标检测,我们的DEA-Net超过了3.08%的平均精度(mAP)。此外,我们的DEA-Net集成了基线的Redet达到了最先进的性能80.37%。在HRSC …
U-NetPlus: A Modified Encoder-Decoder U-Net Architecture for Semantic and Instance Segmentation of Surgical Instrument
Rochester Institute of Technology(罗切斯特理工学院);Center for Imaging Science(影像科学中心);Biomedical Engineering(生物医学工程)
传统的治疗方法由于有限的视野而限制了外科医生的灵活性控制。随着机器人辅助手术的出现,微创手术的医疗技术发生了范式转变。然而,在手术场景中跟踪手术器械的位置是非常具有挑战性的,并且手术工具的准确检测和识别是至关重要的。手术视频帧中基于深度学习的语义分割有可能促进这一任务。在这项工作中,我们修改了名为U-NetPlus的U-Net架构,通过引入预训练的编码器并重新设计解码器部分,通过使用基于最近邻(NN)插值的上采样操作替换转置卷积操作。为了进一步提高性能,我们还采用了非常快速和灵活的数据增强技术。我们在8 x 225帧机器人手术视频序列上训练了框架,这些视频可通过MICCAI 2017 EndoVis Challenge数据集获得,并在8 x 75帧和2 x 300帧视频上进行了测试。使用我们的U-NetPlus架构,我们报告了90.20%的DICE用于二进制分割,76.26%的DICE用于仪器部件分割,46.07%用于仪器类型(即,所有仪器)分割,优于在这些数据上实施和测试的先前技术的结果。
Boundary Discretization and Reliable Classification Network for Temporal Action Detection
School of Computer and Information Engineering, Hefei University of Technology(合肥工业大学计算机与信息工程学院);School of Computer Science and Technology, Hangzhou Dianzi University(杭州电子科技大学计算机科学与技术学院)
PitVQA: Image-grounded Text Embedding LLM for Visual Question Answering in Pituitary Surgery
Wellcome/EPSRC Centre for Interventional and Surgical Sciences (WEISS)(惠康/工程与物理科学研究理事会介入与外科科学中心(WEISS));University College London(伦敦大学学院);The Chinese University of Hong Kong(香港中文大学);National Hospital for Neurology and Neurosurgery(国家神经学与神经外科医院)
Transferable Adversarial Examples for Anchor Free Object Detection
Chengdu University of Information Technology(成都信息工程大学);Keya Medical(科亚医疗);University at Buffalo, State University of New York(纽约州立大学布法罗分校);Microsoft Research Asia(微软亚洲研究院)
ReCal-Net: Joint Region-Channel-Wise Calibrated Network for Semantic Segmentation in Cataract Surgery Videos
Department of Information Technology, Alpen-Adria-Universität Klagenfurt(克恩滕州阿尔卑斯-亚得里亚大学信息技术系);Department of Ophthalmology, Klinikum Klagenfurt(克拉根福医院眼科)
Deep-Learning-Assisted Analysis of Cataract Surgery Videos
Alpen-Adria-Universität Klagenfurt(阿尔卑斯-亚得里亚大学克拉根福分校);University of Applied Sciences Western Switzerland(瑞士西部应用科学大学);University of Bern(伯尔尼大学);University of Geneva(日内瓦大学)