@MASTERSTHESIS\{IMM2015-06865,
    author       = "L. Sommer",
    title        = "Statistical analysis of calcium signaling in transgenic neurons",
    year         = "2015",
    school       = "Technical University of Denmark, Department of Applied Mathematics and Computer Science",
    address      = "Richard Petersens Plads, Building 324, {DK-}2800 Kgs. Lyngby, Denmark, compute@compute.dtu.dk",
    type         = "",
    note         = "{DTU} Compute \& Lundbeck A/S supervisors: Line Katrine Harder Clemmensen, lkhc@dtu.dk, {DTU} Compute, and Ib Vestergaard Klewe, Lars Arvastson. Thesis not publicly available.",
    url          = "http://www.compute.dtu.dk/English.aspx",
    abstract     = "World wide many people suffer from the mental disease schizophrenia, in Denmark alone 25,000 people are diagnosed. Symptoms of the disease include hearing voices that are not there and feeling one is being followed. Existing medical treatment can reduce some of the symptoms, but in many cases, there is severe side effects from the treatment. Despite pronounced side effects the pharmacological treatment of schizophrenia has not changed significantly during the past 60 years and a new treatment is therefore highly desirable.
A newly established cellular model of schizophrenia forms the basis of two image data sets, which are generated and supplied by H. Lundbeck A/S for this thesis. Both data sets are based on calcium imaging of neuronal activity in vitro, but differ in spatial and temporal resolution.
The two supplied data sets are used to gain knowledge about schizophrenia and classifying neurons as wild type or transgenic based on their calcium oscillation patterns. New knowledge about schizophrenia will, hopefully, in the long term lead to an improved treatment of schizophrenia. A network approach is applied to detect structural connectivity differences between the two neuron types and the classification is based on a statistical approach with classifiers such as lasso regularized logistic regression.
An improved discrimination between wild type and transgenic network oscillations has been achieved in this thesis using image data. Several important features, extracted from the image data, that differ between wild type and transgenic neurons have been identified. Further, structural connectivity differences between the two neuron types have been discovered by using a network approach. Based on this work, it can be concluded that Lundbeck can gain more knowledge about schizophrenia by using images of calcium signaling in transgenic neurons instead of the current one dimensional time traces."
}