Image analysis of functional magnetic resonance imaging
Journal
Biomedical Engineering - Applications, Basis and Communications
Journal Volume
13
Journal Issue
5
Pages
248-255
Date Issued
2001
Author(s)
Abstract
Functional MRI can provide important new insights into the phenomenon of brain activity through improved temporal and spatial resolution as well as easiness and safety on repeated or prolonged examination. A revolutionary renewal on the knowledge of functional anatomy of the human brain based on the pre-fMRI technology is thus possible. However, the activity in the brain is so complex that the simplified model-dependent analysis methods may not be sufficient. Therefore, we have endeavored to establish data-driven tools and methodologies for fMRI. Insupervised algorithms based on self-organized artificial neural network and fuzzy logic are used to solve the problem of biased result or misled interpretation frequently encountered in statistical methods depending on prior knowledge of imposed experimental paradigm or assumed pattern of brain response to activation. With these new analysis methods, more accurate activation regions and physiological dynamics can be identified that makes further investigation possible and easier.
Publisher
Institute of Biomedical Engineering
Type
journal article