High-throughput biological imaging promises to be a powerful tool for modern biological research. These imaging technologies are used to monitor the behavior of large populations of cells in response to different experimental conditions. The resulting knowledge can help in drug discovery and in understanding biological phenomena. Different imaging modalities are used for high-throughput imaging. These include 2D fluorescence images, 3D confocal datasets, time-lapse image sequences, and multispectral images. Integrated analysis of these multi-modality spatial, temporal, and spectral data sets for extracting quantitative knowledge is challenging and requires new modeling and data processing tools. This paper presents a multi-layered architecture and spatio-temporal models for analysis of such data. The analysis is divided into low-level and high-level processing. At the lower level, issues like segmentation, tracking and object recognition are addressed, and at the high level, finite state machine- and Petri-net–based models are used for spatio-temporal event recognition. This approach provides a mechanism for extracting high-level spatio-temporal knowledge, and improves searching and retrieval of high-throughput biological imaging data by means of semantic and conceptual queries.
Date of this Version