The University of Edinburgh -
Division of Informatics
Forrest Hill & 80 South Bridge

PhD Thesis #9701

Title:Model Based System for Automated Analysis of Biomedical Images
Date: 1997
Abstract:Today, digital images are used routinely in many areas of biomedical research and clinical practice. The need for reliable quantitative and qualitative analysis of very large numbers of these images has produced an increasing interest in automated Biomedical image processing systems. Traditional image analysis methods (TIA) are built to exploit and take advantage of image data properties, as a result they tend to be fragile with respect to any changes in those properties. For this reason, TIA have been found to be ineffective for the recognition and classification of the inherently irregular and variable biological objects. A potentially more robust and reliable alternative are the Model-based methods as an underlying idea is to use previously identified and explicitly represented properties of the image data in order to establish the best possible interpretation.This thesis presents the development of a probablistic formulation of model-based vision using generalised flexible template models. It includes the design and implementation of a system which extends flexible template models to include grey information in the object representation for image interpretation. This system was designed to deal with microscope images where the different stain and illumination conditions during the image acquisition process produce a strong correlation between density profile and geometric shape.

[Search These Pages] [DAI Home Page] [Comment]