@MASTERSTHESIS\{IMM2007-05391,
    author       = "S. Fischer and P. S. Nielsen",
    title        = "Neuromorphological interpretation of clinical outcome",
    year         = "2007",
    school       = "Informatics and Mathematical Modelling, Technical University of Denmark, {DTU}",
    address      = "Richard Petersens Plads, Building 321, {DK-}2800 Kgs. Lyngby",
    type         = "",
    note         = "Supervised by Assoc. Prof. Rasmus Larsen, {IMM,} {DTU}.",
    url          = "http://www2.compute.dtu.dk/pubdb/pubs/5391-full.html",
    abstract     = "The purpose of this project is to investigate the possibility of analyzing the image data from a pan-European study of the elderly using neuromorphometry. A description of the prevalent methods is given and their individual advantages and disadvantages are analyzed. On the basis of this analysis a relatively new method, tensor-based morphometry (TBM), which still has many unexplored properties, is selected for further investigation.

To facilitate {TBM} analysis two frameworks are suggested — one for crosssectional analysis and one for longitudinal analysis. Both frameworks share certain components: Registration, metric extraction, and statistical analysis.

The properties of the registration that are desirable for respectively the crosssectional and longitudinal studies are analyzed and an image registration is applied.

Metrics for analyzing the registration results are described and the properties of two metrics are investigated and the usefulness in different settings evaluated.

It is shown that some group differences are detectable using the described methods. The metrics are correlated with clinical data both cross-sectionally and longitudinally.

Through the results it is concluded that the proposed framework of methods is viable for neuromorphometry. Through the framework it is possible to detect both general atrophy and pathologies in single subjects and differences between subjects which can be correlated with clinical parameters."
}