Reconstructing cell lineages from 3D time-lapse microscopy requires distinguishing cell motion from division and inferring temporally consistent connections. When only a subset of cells is annotated, however, training procedures that treat unannotated regions as background or incorrect correspondences can be problematic. We propose a cell tracking pipeline that integrates detection from sparse annotations, calibration of candidate connections using image features, and joint selection of lineage graph events. For detection, we combine the outputs of 2.5D and 3D models trained to suppress erroneous background supervision at unannotated cells. We then calibrate ordinary link and division probabilities using motion features from dense flow, correspondence features from a learned cell-matching model, and features from models that identify cells immediately before division, dividing parents, and daughters immediately after division. Ordinary links, divisions represented as daughter pairs, and node inclusion are selected jointly under a surrogate objective derived from the evaluation metric, followed by gap filling and temporal coordinate refinement. The pipeline achieves an overall score of 0.9762 in 5-fold cross-validation on 199 videos from 2 zebrafish embryos. With candidates and calibrated values held fixed, joint selection outperforms a sequential method that allows replacement of existing divisions and reoptimization of ordinary links by 0.0105. The improvement arises primarily from division recovery.
Cell Behavior Video Classification Challenge, a benchmark for computer vision methods in time-lapse microscopy
Euler Institute, Faculty of Informatics, Università della Svizzera italiana(欧拉研究所,信息学院,瑞士大学);International Center for Advanced Computing in Medicine (ICAM), University of Pavia(国际医学先进计算中心(ICAM),帕维亚大学);Imaging Platform, ACRF INCITe Centre, Garvan Institute of Medical Research(成像平台,ACRF INCITe中心,嘉文医学研究所);Tacoma School of Engineering & Technology, University of Washington(塔科马工程与技术学院,华盛顿大学);Data Science, Institute for Computing and Information Sciences, Radboud University(数据科学,计算与信息科学研究所,拉德堡德大学);Imaging Core, Lerner Research Institute, Cleveland Clinic(成像核心,勒纳研究研究所,克利夫兰诊所);Institute of Biomedical Engineering, University of Toronto(生物医学工程研究所,多伦多大学);Center for Research in Computer Vision, University of Central Florida(计算机视觉研究中心,佛罗里达大学);Computational Biology Group, Data Science Platform, Garvan Institute of Medical Research(计算生物学小组,数据科学平台,嘉文医学研究所)
Tracking Intermittent Particles with Self-Learned Visual Features
Institut Pasteur(巴斯德研究所);CNRS(法国国家科学研究中心);LTCI, Telecom Paris, Institut Polytechnique de Paris(巴黎理工学院巴黎电信学院LTCI);Department of Biological Sciences, Columbia University(哥伦比亚大学生物科学系)
UmUTracker: A versatile MATLAB program for automated particle tracking of 2D light microscopy or 3D digital holography data
Department of Physics, Umeå University(于默奥大学物理系)
我们提出了一个通用的和快速的MATLAB程序(UmUTracker),自动检测和跟踪颗粒通过分析视频序列获得的光学显微镜或数字在线全息显微镜。我们的程序检测粒子的二维横向位置与等腰三角形变换的基础上的算法,并重建其三维轴向位置的瑞利-索末菲模型使用径向强度分布的快速实现。为了验证我们的程序的准确性和性能,我们首先使用明场和数字全息显微镜跟踪聚苯乙烯颗粒的2D位置。其次,我们通过分析合成和实验获得的全息图来确定3D粒子位置。最后,为了突出完整的程序功能,我们在100微米的高流量室中对微流体流动进行了分析。这一结果与计算流体动力学模拟一致。在普通的台式计算机上,UmUTracker可以在1024 x 1024的图像中以每秒5帧的速度检测、分析和跟踪多个粒子,模板大小为201 x 201。为了增强可用性,并使其易于实现新的功能,我们使用面向对象的编程。UmUTracker适用于与以下相关的研究:粒子动力学、细胞定位、胶体和微流体流动测量。
Trackastra: Transformer-based cell tracking for live-cell microscopy
Institute of Bioengineering, School of Life Sciences, EPFL(EPFL生命科学学院生物工程研究所);Center for Scalable Data Analytics and AI (ScaDS.AI)(可扩展数据分析与人工智能中心(ScaDS.AI));Faculty of Computer Science, TU Dresden(德累斯顿工业大学计算机科学学院)
Counting Cells in Time-Lapse Microscopy using Deep Neural Networks
Istituto Sistemi e Electronica Applicata (ISEA)(应用系统与电子研究所);Scuola universitaria professionale della Svizzera italiana (SUPSI)(瑞士意大利语区应用科技大学);Grupo de investigacion SISTEMIC(SISTEMIC研究组);Facultad de Ingenieria, Universidad de Antioquia UdeA(安蒂奥基亚大学工程学院)
Accurate Tracking of Arabidopsis Root Cortex Cell Nuclei in 3D Time-Lapse Microscopy Images Based on Genetic Algorithm
Graduate School of Science and Technology, Nara Institute of Science and Technology(奈良科学技术大学研究生院);Bioresource Engineering Laboratory, Ishikawa Prefectural University(石川县立大学生物资源工程实验室);Department of Engineering Informatics, Osaka Electro-Communication University(大阪电讯大学工程信息学系);Exploratory Research Center on Life and Living Systems, National Institutes of Natural Sciences(国家自然科学研究院生命与生活系统探索研究中心);College of Computer Science and Technology, Zhejiang University(浙江大学计算机科学与技术学院)
SPF-CellTracker: Tracking multiple cells with strongly-correlated moves using a spatial particle filter
Kanazawa University(金泽大学);Kyushu Institute of Technology(九州工业大学);University of Tokyo(东京大学);Kyushu University(九州大学);Institute of Statistical Mathematics(统计数理研究所);JST-CREST(日本科学技术振兴机构CREST)
PyUAT: Open-source Python framework for efficient and scalable cell tracking
Institute of Bio- and Geosciences, IBG-1: Biotechnology, Forschungszentrum Jülich(于利希研究中心生物与地球科学研究所IBG-1生物技术分部);Computational Systems Biotechnology (AVT.CSB), RWTH Aachen University(亚琛工业大学计算系统生物技术研究所(AVT.CSB))
Recursive Deep Prior Video: a Super Resolution algorithm for Time-Lapse Microscopy of organ-on-chip experiments
University of Bologna(博洛尼亚大学);University of Tor Vergata(罗马第二大学(托尔·韦尔加塔大学));Institute Curie Centre de Recherche(居里研究所研究中心);Paris Sciences et Lettres Research University(巴黎文理研究大学)
DistNet: Deep Tracking by displacement regression: application to bacteria growing in the Mother Machine
Laboratoire Jean-Perrin UMR 8237, Sorbonne Université(索邦大学让-佩兰实验室);SABILAb;CMAP, Ecole Polytechnique, Université Paris-Saclay(巴黎萨克雷大学综合理工学院应用数学中心);Heuritech
母机是一种流行的微流体设备,可以通过显微镜对数千个细胞进行长期的延时成像。它已成为单细胞水平定量分析和表征许多细胞过程,如基因表达和调控,诱变或对抗生素的反应的有价值的工具。对这些实验产生的大量数据进行自动化和定量分析现在是限制性的一步。特别是相衬显微镜成像的细菌细胞的分割和跟踪--与高通量数据兼容的错误率--是一个具有挑战性的问题。在这项工作中,我们描述了一种新的配方的多目标跟踪问题,其中跟踪是通过回归的细菌的位移,允许同时跟踪多个细菌,尽管他们的生长和分裂随着时间的推移。我们的方法执行联合分割和跟踪,利用顺序信息,以提高分割精度。我们引入了一个深度神经网络架构,利用自注意机制,产生极低的跟踪错误率和分割错误率。与最先进的方法相比,我们展示了卓越的性能和速度。我们的方法被命名为DistNet,代表DISTance+Displacement Segmentation and Tracking Network。虽然这种方法是特别适合于母机显微镜数据,其一般的联合跟踪和分割配方可以应用于许多其他问题与不同的几何形状。
TransientTrack: Advanced Multi-Object Tracking and Classification of Cancer Cells with Transient Fluorescent Signals
Institute for Biomedical Informatics, Faculty of Medicine;University Hospital Cologne, University of Cologne, Cologne, Germany.;Faculty of Mathematics;Natural Sciences, University of Cologne, Cologne, Germany.;Center for Molecular Medicine Cologne (CMMC), Faculty of Medicine;Cologne Excellence Cluster on Cellular Stress Responses in Aging-Associated Diseases (CECAD), University of Cologne, Cologne, Germany;Charité Comprehensive Cancer Center, Charité Universitätsmedizin, Berlin, Germany.;Centre for Genomic Regulation, Barcelona, Spain.
CellTrack R-CNN: A Novel End-To-End Deep Neural Network for Cell Segmentation and Tracking in Microscopy Images
University of Sydney(悉尼大学);University of New South Wales(新南威尔士大学);Brigham and Women’s Hospital(布莱根妇女医院);Harvard Medical School(哈佛医学院);Sydney Nano Institute(悉尼纳米研究所)
Motion-guided sparse correction enables expert-quality point tracking across diverse microscopy regimes
Department of Computer Science, Old Dominion University(奥德赛大学计算机科学系);Department of Biology, Massachusetts Institute of Technology(麻省理工学院生物学系);The Picower Institute for Learning and Memory, Massachusetts Institute of Technology(麻省理工学院学习与记忆研究所);Department of Physics and Technology, UiT--The Arctic University of Norway(挪威北极大学物理与技术系);Department of Physics, University of Oslo(奥斯陆大学物理系);School of Data Science, Old Dominion University(奥德赛大学数据科学学院);Department of Physics, Old Dominion University(奥德赛大学物理系)
Automated Measurements of Key Morphological Features of Human Embryos for IVF
Department of Molecular and Cellular Biology(分子与细胞生物学系);School of Engineering and Applied Sciences(工程与应用科学学院);Tel Aviv Sourasky Medical Center(特拉维夫苏拉斯基医疗中心);Harvard Graduate Program in Biophysics(哈佛大学生物物理学研究生项目);Harvard University(哈佛大学)
Unsupervised cell segmentation by fast Gaussian Processes
Department of Statistics and Applied Probability, University of California, Santa Barbara, CA(统计与应用概率系,加州大学圣芭芭拉分校);Department of Electrical and Computer Engineering, University of California, Santa Barbara, CA(电气与计算机工程系,加州大学圣芭芭拉分校)
Early-detection and classification of live bacteria using time-lapse coherent imaging and deep learning
University of California, Los Angeles(加利福尼亚大学洛杉矶分校);California NanoSystems Institute(加州纳米系统研究所);Istanbul University(伊斯坦布尔大学);David Geffen School of Medicine(大卫·格芬医学院)
我们提出了一种计算活细菌检测系统,该系统定期捕获60 mm直径琼脂板内细菌生长的相干显微镜图像,并使用深度神经网络分析这些时间流逝的全息图,以快速检测细菌生长和相应物种的分类。我们的系统的性能通过快速检测大肠杆菌和总大肠菌群(即,产气克雷伯氏菌和肺炎克雷伯氏菌肺炎亚种。pneumoniae)。这些结果与基于金标准培养的结果进行了对比,与环境保护局(EPA)批准的分析方法相比,细菌生长的检测时间缩短了>12小时。我们的实验进一步证实,该方法在7-10 h内成功检测90%的细菌菌落(12 h内>95%),精确度为99.2- 100%,并在7.6-12 h内正确鉴定其种类,准确率为80%。使用生长培养基中的样品预孵育,我们的系统在总测试时间的9小时内实现了~1个菌落形成单位(CFU)/L的检测限(LOD)。这种计算细菌检测和分类平台具有很高的成本效益(每次测试约0.6美元)和高通量,整个平板表面的扫描速度为24 cm 2/min,非常适合与目前用于琼脂平板细菌检测的现有分析方法集成。在深度学习的支持下,这种自动化且具有成本效益的活细菌检测平台可以通过显着减少检测时间,自动识别菌落,而无需标记或专家,从而为微生物学中的广泛应用带来变革。
Deep Joint Denoising and Detection for Enhanced Intracellular Particle Analysis
Erasmus University Medical Center(伊拉斯姆斯大学医学中心);Utrecht University(乌得勒支大学);University of New South Wales (UNSW)(新南威尔士大学);NVIDIA Corporation(英伟达公司);SURFsara