As we move into the future, biomedical science is becoming more and more data-driven. The vast amounts of information that are amassed in many areas of the field are becoming increasingly more difficult to manage due to the sheer size of the data sets and the fact that it is becoming more visually oriented. This proposal has selected one area, confocal microscopy, where the large amounts of visual data are produced at a rapid pace with little means to organize it into a usable format for others to access. It is also apparent that with the changes in information technology, the means to collaborate on such information via high speed networks has compounded the problem of sharing large image data sets across the Internet. Because imaging and visualization are technologies which consume large amounts of both computational and networking resources, it seemed logical to work with researchers collaborating in an image intensive domain. The University of Michigan is undertaking to examine the means by which to coordinate image data, organize, classify, retrieve and transmit the large data sets that are being produced in the area of confocal microscopy. To do this, a national repository is needed to store this ever increasing information into a form that is accessible to all those researchers in the field around the nation. This is significant because it will provide a biomedical group a number of advantages: (1) a seamless distributed storage and retrieval system; (2) a shared, distributed resource; (3) easy, transparent access and (4) shared input capability by scientists. The repository is also needed to solve many of the issues that currently face those in biomedicine that seek to use this information for research, education and clinical purposes. It would address the following issues such as the growth of the data, problems of retrieval and the specific problems of computation and communication of an image repository. The purpose of this project is to develop strategies that support collaborative research which depend on the availability of high performance computing and networking technologies. Specifically, we plan to (1) develop a network-based image repository for sharing of the data from confocal microscopy including the ability to automatically convert images into the format(s) most appropriate for the users; (2) examine the information retrieval factors which will influence the use of large image repositories with the long-term goal of constructing a classification and/or indexing model that can be generalized to other biomedical domains that deal with large amounts of visual data; (3) examine the parameters influencing the organization, storage, retrieval, and transmission of large data sets which require high power computation and communication via high bandwidth across the Internet and (40 organize the national cooperative group in confocal microscopy to provide cohesiveness and focus for this project and to better serve a community of committed researchers.
