Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.540495
Title: The use of statistical parameter mapping for longitudinal PET/CT studies of mouse brain metabolism
Author: Awikunprasert, Panatsada
Awarding Body: University of Aberdeen
Current Institution: University of Aberdeen
Date of Award: 2010
Availability of Full Text:
Access through EThOS:
Full text unavailable from EThOS. Please try the link below.
Access through Institution:
Abstract:
In this work, we aimed to assess the possibility of applying the SPM method to a smallanimal PET data set including the use of the SPM registration option and the investigation of its applications in translational studies. We also evaluated and determined which image registration technique is best suited for aligning mice brain PET/CT images. Methods: The data were acquired from a cohort of C57BL/6 wild-type mice which underwent PET/CT scanning, producing the matched FDG-PET and CT brain images. Three different registration techniques to align mouse brain PET images were investigated: 1) PET/PET template 2) CT/CT template and 3) the segmented mask brain registration (described below). The automatic registration package within SPM was used to register the mice brain images. Optimal parameter settings for registering mice brain images were examined. Multiple measures of accuracy including: visual inspection, volume overlap, distance error and mutual information values were used to evaluate the image registration techniques. Results: In comparison between the PET-PET and CT-CT methods, there were no statistically significant in the means of the volume overlap measures. However, the mean distance error from the CT-CT method was significantly lower than for the PET/PET method. The qualitative visual inspection results showed that there were three registered images from the PET-PET method, and two from the CT-CT method were misalignment could be detected whereas there was no image misalignment detected from the mask-mask technique. Conclusion: The use of the segmented (mask) brain registration method produced the most successful results. A major benefit of including a mask brain in the registration is that it excludes the non-brain regions which contain unwanted signal (i.e. bright intensity of eyes), ensuring that the registration procedure used specifically the brain area for registering, avoiding the other bright areas of non-brain structures.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.540495  DOI: Not available
Keywords: Brain
Share: