[diss] Perustieteiden korkeakoulu / SCI

Permanent URI for this collectionhttps://aaltodoc.aalto.fi/handle/123456789/52

Browse

Browsing [diss] Perustieteiden korkeakoulu / SCI by Degree programme/Major subject "Sovellettu matematiikka"

Filter results by typing the first few letters
Now showing 1 - 1 of 1
  • Quality improvements for multi-modal neuroimaging
    (2007-09-17) Seppä, Mika
     Doctoral dissertation (article-based)
    This work, concentrating on improving the quality of multi-modal medical image fusion, was carried out in the Brain Research Unit of the Low Temperature Laboratory at Helsinki University of Technology (TKK) in close collaboration with the Advanced Magnetic Imaging (AMI) Centre of TKK. Modern medical imaging devices produce large amounts of highly detailed information about the anatomy and function of various body parts. Different imaging modalities are typically sensitive to different properties of the underlying tissue and therefore produce complementary information. In multi-modal neuroimaging, data from different modalities is fused together to facilitate better analysis of the structure and activity of the brain or of other parts of the nervous system. Multi-modal image fusion in human neuroimaging has many uses both in clinical settings and in research. Since the different imaging modalities reveal different properties of the nervous system, joint visualization helps to combine all this information for interpretation. Typically, data are combined to visualize the anatomical locations of the functional activations. In addition to joint visualization, multi-modal neuroimaging can also incorporate information from one modality to the analysis of data in another modality and thereby lead to more accurate results. In the core of multi-modal imaging lies image registration that brings together the information from two or more imaging modalities. Since the spatial alignments and resolutions of the registered images typically differ, resampling is required to bring the data into a common coordinate frame for visualization or for further analysis. This thesis work concentrates on the three key stages of the multi-modal image fusion: registration, resampling, and visualization. The introduced enhancements for mutual-information registration allow for sub-sample accuracy even in the worst-case scenarios and the novel two-stage resampling algorithm produces smaller resampling errors than any of the currently used methods. Furthermore, the proposed enhancement to combine triangle meshes with volume rendering techniques provides fast high-quality visualization. In addition to these technical improvements, an application of diffusion tensor imaging to delineate the course of peripheral nerves is presented.