Enhanced Cuckoo Search Algorithm With Metaheuristic Components For Extracting The Maxima Of The Orientation Distribution Function
Loading...
Date
2018-06
Authors
Shehab, Mohammad Mohammad Said
Journal Title
Journal ISSN
Volume Title
Publisher
Universiti Sains Malaysia
Abstract
The Diffusion-Weighted Magnetic Resonance Imaging(DW-MRI) is a promising method
for non-invasive investigation of anatomical connectivity in the human brain. The raw data
acquired from the MRI scanner may not be directly usable by the specialists. Therefore, new
methods are required to make more reasonable representations of the data to extract the required
information from them. The initial representation of the MRI data is the huge groups of
fibers. These fibers contain fiber crossing bundles, which link the functional brain areas all
together as a complex net-work of neural fiber tracts. Q-ball imaging (QBI) is a Diffusion MRI
reconstruction technique which has been proven very successful in resolving multiple
intravoxel fiber orientations in MRI (i.e., fiber crossing) based on the standard computation of
the Orientation Distribution Function (ODF), which is a 3- Dimension spherical function
founded to detect the dominant fiber orientations in the underlying volume of a pixel (voxel).
This dissertation presents a new method to solve ODF problem through adapting one of the
metaheuristic algorithms, namely, Cuckoo Search Algorithm (CSA) for the ODF. The adaptation
involved preparing the synthetic data for testing. The results were within the range of
previous work and better comparing the other algorithms. However, some shortcomings in
the convergence rate and local exploitation were determined and addressed by enhancing with
known metaheuristic components. Three successive enhancement versions are proposed
tions of various categories. This is to test each version before using it for the ODF. The second
step compares against five other methods using synthetic data. The ODFs reconstructed by
CSAHC-ODF are sharper and more accurate ODFs than the original image and extracts more
accurate maxima.
Description
Keywords
Cuckoo search algorithm with metaheuristic components , maxima of the orientation distribution function