Robust Characterization of Moving Objects for Subcellular Time-lapse Assays

描述(申请人提供)：新一代显微镜和荧光探针技术使定量描述活细胞中离散蛋白质和细胞器的动态成为可能。尽管这些分析越来越普遍，但最先进的图像信息学工具使用过时的检测和跟踪技术，这些技术最初并不是为具有挑战性的亚细胞环境而设计的，无法提供足够的稳健性、准确性和自动化。因此，对这类数据的定量动力学分析在很大程度上仍然是手工进行的；这是乏味、耗时和主观的。缺乏合适的亚细胞延时分析表征工具已成为细胞生物学和疾病研究的关键瓶颈。我们正在开发下一代显微图像分析产品SVCell，专注于最具挑战性的应用中的信息学创新，包括亚细胞和活细胞，时间推移分析。我们建议扩展SVCell的技术，以便在时间推移分析中对多个亚细胞运动对象进行稳健、准确和高度自动化的动力学表征。具体目标是：1)完成生物目标的自适应时空目标检测的开发；2)完成时移显微图像序列中鲁棒跟踪的开发；3)使用多个数据集定量评估自适应目标检测和稳健跟踪性能在集成的SVCell动态阿尔法中的性能。该方案的创新之处包括利用完整图像序列的稳健跟踪算法和自适应弹性目标检测。SVCell Kinetics将是我们将向广大生命科学界销售的一套SVCell图像信息学工具中的一种。这一工具可能对艾滋病毒-1、人乳头瘤病毒、单纯疱疹病毒、腺病毒、流感和其他病毒，以及糖尿病、癌症和神经退行性疾病相关的研究，如阿尔茨海默氏症、亨廷顿病和帕金森病以及艾滋病的综合症状产生特别的影响。它还可以提高药物开发中靶点发现的效率，并加强基于病毒载体的基因递送系统的研究和生产。

DESCRIPTION (provided by applicant): A new generation of microscope and fluorescent probe technologies are enabling the quantitative characterization of the dynamics of discreet proteins and organelles in living cells. Though these assays are increasingly common, the state-of-the-art image informatics tools, which use outdated detection and tracking technologies not originally designed for the challenging subcellular environment, fail to provide adequate robustness, accuracy and automation. Consequently, quantitative kinetic analysis of such data remains largely manual; which is tedious, time consuming and subjective. The lack of an adequate subcellular time-lapse assay characterization tool has become a critical bottleneck in cell biology and disease research. We are developing a next generation microscopy image analysis product, SVCell, focusing on informatics innovation in the most challenging applications including subcellular and live cell, time-lapse analyses. We propose to extend SVCell's technology for robust, accurate and highly automated kinetic characterization of multiple subcellular moving objects in time-lapse assays. The specific aims are: 1) Complete the development of the adaptive spatio-temporal object detection of biological objects; 2) Complete the development of the robust tracking in time-lapse microscopy image sequences; and 3) Quantitatively evaluate adaptive object detection and robust tracking performance in an integrated SVCell Kinetic alpha using multiple data sets. The innovation of this proposal includes robust tracking algorithms utilizing the complete image sequence and adaptive elastic object detection. SVCell Kinetics will be 1 in a suite of SVCell image informatics tools that we will market to the broad life sciences community. This tool could have particular impact on HIV-1, HPV, herpes simplex, adenovirus, influenza and other viruses, as well as diabetes, cancer and neurodegenerative disease related research such as in Alzheimer's, Huntington's and Parkinson's' disease and the AIDs dimentia complex. It could also increase the efficiency of target discovery in drug development, and enhance viral vector based gene delivery systems research and production.