Tutorial I: Overview of fMRI Analysis

Purpose and aims

This tutorial provides an introduction to the key theoretical concepts behind the analysis of functional magnetic resonance imaging (fMRI) time-series data. Functional MRI is a non-invasive brain imaging technique used by clinicians, neuroscientists and commercial parties to answer questions about the functioning of the brain. Functional MRI research often involves an experimental task that is carried out by a subject inside the MRI scanner with the aim to engage a set of predicted brain circuits, based on the cognitive requirements of this task. The activity within brain circuits can be measured reliably using fMRI, because a coupling between neural activity and blood oxygen consumption exists and its mechanism is increasingly understood. It is this indirect measure of neural activity, called the haemodynamic response that lies at the heart of this imaging technique. After the data has been gathered, it is processed in a series of image processing steps, that transforms the raw k-space data into real images and finally into statistical maps, also known as activation images. Understanding the purpose and meaning of the image processing steps that transform unprocessed data into statistical maps lies at the very basis of examining neuronal function with the help of fMRI. In this tutorial, the image processing steps are introduced that are part of a commonly used imaging processing pipeline. Note that the emphasis in this tutorial is put on generating statistical maps from unprocessed data; MRI physics and neurovascular coupling are introduced in separate tutorials on MRI-TUTORIAL.COM.