Dynamic Resting-State fMRI Analysis Methods

Dynamic resting-state fMRI is a method that measures the blood-oxygen-level-dependent (BOLD) signal, as a result of neurovascular coupling, while the brain is at rest. With this, the mesoscale brain network functional connectivity can be assessed. Typically, traditional measures use an average of all repetitions (volumes) to quantify the connectivity between two predefined regions. However, novel dynamic functional connectivity measures are able to assess the network time-dependent relationship by using all repetitions. Two such promising methods are the quasi-periodic patterns (QPPs) and co-activation patterns (CAPs) which depict whole brain dynamic states. The QPPs are a sliding window analysis method which means that these patterns have a predefined temporal length (set number of volumes) that is used as a template to detect whole brain recurring patterns in the brain. On the other side, CAPs are a frame-wise method, which defines each volume as a CAP and further uses k-means clustering to define the most prominent CAPs that govern the brain at rest.