fMRI Imaging Based Human Brain Parcellation Methods: A review

• Published in: 2020 5th International Conference on Advanced Technologies for Signal and Image Processing (ATSIP) pp. 1 - 5
• Main Authors: Abdedayem, Fatma, Kallel, Fathi, Chaabane, Marwa, Hamida, Ahmed Ben, Sellami, Lamia
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-09-2020
• Subjects: correlation matrix; fMRI; regions of interest (ROIs)
• ISSN: 2687-878X
• DOI: 10.1109/ATSIP49331.2020.9231946

In the context of functional magnetic resonance imaging (fMRI) images, human brain parcellation is a very critical issue for human brain network generation, analysis and functional connectivity researches. Thus, brain organization has been defined with specific topographies at different scales where brain is dividing into different areas or regions which are interconnecting closely between them and each one is represented by a node with specific local properties.Typically, a huge number of recent parcellation studies prove the important role of fMRI based approaches to parcelate the brain into different interest regions by using several distinct generated Atlases.In this paper, we present a comparison between different Atlases that represent the map of parcellation. More specifically, we study the initial steps which are applied later with graph theories that has been used in a huge number of previous works in order to detect and describe neurodegenerative diseases. In our work, we have started by calculating the mean correlation matrix using distinct Atlases. We have chosen popular Atlas such as probabilistic Atlas, YEO Atlas, Power Atlas, Crad Atlas, Fair Atlas and Dos Atlas.Our experimental results are based on testing fMRI images taken from the famous database for Alzheimer named ADNI.