A DISTRIBUTED IMAGING FRAMEWORK TO EXTRACT, ANALYSE AND VISUALIZE CELL FEATURES IN MULTI-DIMENSIONAL BIO-IMAGE DATASETS

摘要

Cunrent imaging technologies can generate datasets of immense size and new image processing applications are applying ever more complex algorithms to these datasets. Parallel computing resources are commonly employed to obtain results in reasonable timescales. This paper outlines the development of a distributed framework to provide support for concurrent processing across a parallel system, and to apply this framework to the 3D image processing and visualization domains. Specifically, image processing to identify and extract cell image features in large multi-dimensional bio-image datasets, and visualization of the results. A novel aspect of this integrated framework is to allow feature extraction on the fly for the analysis of spatio-temporal events and their immediate visualisation, which greatly enhances the interpretation of the data.