Investigation and implementation of image processing algorithms for medical applications

Abstract

Functional Magnetic Resonance Imaging (fMRI) allows non-invasive assessment of

brain activity using MRI signal changes related to functional brain activity. The main

aim of fMRI analysis is to detect changes in blood oxygenation and flow that occur

when in response to neural activity. Activation maps can be produced using fMRI

which displays parts of the brain that are involved in a particular mental process.

This project aims to investigate fMRI data analysis techniques for activated voxel

detection. The techniques include Correlation Coefficient, General Linear Model and

Mutual Information. The project uses Statistical Parametric Mapping (SPM) and

Receiver Operating Characteristic (ROC) curves to compare the effectiveness of the

three mentioned fMRI data analysis techniques.