Main page
Author's kit
Registration
Program
Organization
Venue
More information
wcsb06@cs.tut.fi
Hosted by
TUT CSB group
Abstract --- Antti Niemistö, Institute of Signal Processing, Tampere University of Technology, FinlandAutomated Quantitative Analysis of Biomedical Microscopy Images Traditionally biological samples have been analyzed manually by visual inspection under the microscope. For example, a simple task could be to count the number of cells in a cell population. Analysis performed in this way is naturally very labor intensive, tedious, and slow. Moreover, if a quantitative manual analysis is made by two different persons, the results may not be the same. This is known as inter-observer variability. Intra-observer variability can also be observed, that is, when the same person performs the analysis twice, the results may differ. If a digital camera is attached to the microscope and digital images of the biological samples are obtained, automated image analysis can be used to overcome all of the above mentioned problems. The same criteria and algorithms are always used in detecting and quantifying the desired features from the images, and the analysis is always performed objectively. Since the analysis can always be performed in exactly the same way, results obtained with automated image analysis are also reproducible. Although the solution to an image analysis problem depends heavily on the nature of the image data, the general image analysis procedure is usually the same. The main steps of the procedure are image acquisition, image pre-processing, image segmentation, feature extraction, validation, and data analysis. In the talk, these steps are described with the aid of illustrative example images. |