Collaborative Systems for Analyzing Tissue Microarrays

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Tissue microarray technology holds great potential for reducing the time and cost associated with conducting research in tissue banking, proteomics, and outcome studies. However, capturing, organizing, updating, exchanging, and analyzing the data generated by this technology creates a number of significant challenges. The sheer volume of data, text, and images arising from even limited studies involving tissue microarrays can over time quickly approach those of a small clinical department. This challenge is compounded by the fact that large-scale projects in several areas of modern research often involve multi-institutional efforts in which investigators and resources are spread out over multiple campuses, cities, and states. Future progress in these areas of research will depend on the capacity of individuals from disparate sites to interact, share and assimilate information derived from a variety of sources and modalities. The central objective of this grant proposal is to develop an automated tissue microarray analysis system which features distributed imaging and data management capability in a collaborative environment which supports impromptu telemedicine applications in pathology and oncology. The University of Medicine & Dentistry of New Jersey will collaborate with the Cancer Institute of New Jersey, Rutgers University and the University of Pennsylvania School of Medicine to develop the algorithms, computational tools, and technologies which enable individuals from disparate research and clinical sites to reliably image, analyze, archive and share tissue microarrays, pathology specimens, and correlated data in multi-user environments in order to support collaborative clinical and research activities. The aims of the proposed project are (1) to design, develop, and evaluate a portable suite of software tools for performing unsupervised imaging and archiving of tissue microarrays in multi-user environments; (2) to evaluate and optimize the performance of a set of newly developed color decomposition algorithms for use in quantitative characterization of expression in cancer tissue arrays; (3) to investigate the use of a variable bandwidth mean shift algorithm for segmenting constituent array discs; (4) to investigate the use of a novel algorithm based on color active contour models and robust estimation for use in sub-cellular localization and compare with mean shift results; (5) To investigate the use of texton libraries for characterizing expression patterns in cancer tissue arrays; (6) to develop a Tissue Microarray Repository (TMR) subsystem which enables individuals from disparate research and clinical sites to populate the database with new cases including correlated image metrics and imaged arrays; (7) to integrate the results from (1-5) into a reliable image processing module of the automated TMA analysis system; (8) to deploy the web-enabled TMA analysis server and client to strategic sites throughout UMDNJ, UPenn, and the Cancer Institute and evaluate performance in multi-user, multi institutional environments under real operating conditions; and (7) to refine the system based on performance studies and make web-enabled software and computational tools and supporting documentation available to the clinical and research communities.