Beta 1


Title Automated Characterization and Recognition of 2D and 3D Brain Structure in MRI for Diagnostic Support
Author Blaszczyk, Maciej Jerzy
Supervisor Larsen, Rasmus (Image Analysis and Computer Graphics, Department of Informatics and Mathematical Modeling, Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark)
Institution Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark
Thesis level Master's thesis
Year 2008
Abstract This thesis describes a method for automated characterisation and recognition of Corpus Callosum brain structure from 3-dimensional magnetic resonance images and relation of derived shape parameters to clinical and cognitive performance data. Thesis is a product of project research carried due to a collaboration between Informatics and Mathematical Modelling, DTU and the MR Department at Hvidovre University Hospital. Medical data was gathered during Leukoaraiosis And DiSability in the elderly (LADIS) study. Thesis is focused on accurate characterization of dependence of change in shape of Corpus Callosum during the period of three years for each subject, as well as corelating these changes with changes in cognitive performance data. Prior to characterisation of Corpus Callosum shape, the magnetic resonance data is used to define the most proper Mid Saggital surface in both baseline and follow up data sets, to reveal the corresponding parts of Corpus Callosum. The surface determination is based on local symmetry measures along respective parts of MRI data and Median as well as Average filtering of the results. Shape definition is based on contour defined by means of dynamic programming on image gradient values. Shape changes and variations among the subjects are examined by means of Principal Component Analysis.
Series IMM-M.Sc.-2008-87
Fulltext
Original PDF ep08_87_net.pdf (2.92 MB)
Admin Creation date: 2008-09-03    Update date: 2008-09-03    Source: dtu    ID: 222840    Original MXD