[SOUND] Hi. In this module, we're going to start talking about a new subject which is brain connectivity. So, previously we talked, in the previous course, we talked about data acquisition and reconstruction and pre-processing. Then we started talking about data analysis. And we talked mostly about brain mapping and localizing brain activation. Now we're going to shift our focus towards the assessing brain connectivity. So human brain mapping has primarily been used to construct maps indicating regions of the brain, activated by a certain task. And recently, there's been a great increased interest in augmenting these types of analysis with connectivity studies. This study seek to describe how brain regions interact with one another, and how these interactions depend on experimental conditions and behavioral measures. It has become common practice to talk about brain networks. These are sets of interconnected brain regions with information transfer between the different regions. So to construct a region, we'd first take the brain here and define a set of nodes. These are regions of interest. In this little cartoon, we have three regions of interest, A, B and C. Next, we estimate the set of connection or edges between the nodes. These edges can either be bidirectional or directional. Here we see an example of both. So there's a unidirectional edge between A and B and then, there's a bidirectional edge between B and C. This can also be represented in matrix format, in this so called adjacency matrix. Here we have ones in positions where there's an edge from one node to the other. So between A and B we have a 1, and between B and C, and C and B we have 1's, otherwise they're 0's. This is what's called an adjacency matrix, and this is a representation of matrix format of the network. So a number of methods have been suggested in literature to quantify the relationship between nodes or regions. And their appropriateness depends on a number of factors. These include what type of conclusions was interested in making, what type of assumptions one willing to make, the level of analysis, and the modality used to obtain the data. If have surveyed the brain connectivity literature, you quickly come across two pieces of notation. There's one often talks about functional connectivity and effective connectivity. So functional connectivity is defined as the undirected associated between two or more fMRI time series. We extend that definition to also include performance and physiological variables. The functional connectivity makes statement about the structure of relationships among brain regions, but it usually doesn't make any underlying assumptions about the biology. Here's an example of a simple functional connectivity result. Here we're saying that the region VMPFC is say, correlated with these other three regions. Now that would be a functional connectivity study, the results of a functional connectivity study. Methods for performing functional connectivity include seed analysis and inverse covariance methods which we're going to be talking about in the next module. And multivariate decomposition methods such as PCA or principle, components analysis. ICA or independent component analysis. Or PLS or partial least squares. In this, we will be talking about in the subsequent module. So, a goal of functional connectivity analysis is to make inferences on the structure of relationships among different brain regions. So we want to make conclusions such as these regions form a network. Regions are more connected during task A than task B. Or, this task is associated with activation of pain pathways. The other type of connectivity that we're going to be talking about in later modules, is effective connectivity. And effective connectivity is defined as the directed influence of one brain region on the physiological activity recorded in other brain regions. So this is, makes stronger conclusions. So this claims to make statements about causal effects among tasks and regions. And here in effective connectivity methods, we typically make anatomically motivated assumption and restrict inference to networks comprising of a number of pre-selected regions of interest. Here's an example of a result in effective connectivity. Here we might say that, activation in V1 leads to activation in V5 which leads to activation in PPC. So notice here that we're making directional assumptions, because we're talking about the directed influence on one regions on the physiological activity in other brain regions. Methods for assessing effective connectivity include structural equation modeling, Granger causality. Dynamic cause and modeling, and Bayes nets. And so we'll touch upon all of these methods in later classes. A goal of effective connectivity analysis is to make statements about causal effects among tasks and regions. We might want to say things like, the frontal cortex enhances connectivity between visual areas and hippocampus. Or the VMPFC inhibits the amygdala. Connectivity can be studied at different levels with different interpretations at each. For example, we might study connectivity across time and this can reveal networks that are dynamically activated across time. Connectivity across trials can identify coherent networks of related activations. Connectivity across subjects can reveal patterns of coherent individual differences. And finally, connectivity across studies can reveal tendencies for studies to co-activate within sets of regions. Okay. So that's the end of this module. In the coming modules, we'll be talking about functional. Different methods for assessing functional connectivity. So the next module, we'll talk about seed analysis. And the partial correlation and the like, and then in the following module, we'll talk about ICA and PCA and these multi-varied decomposition methods. Okay, I'll see you then. Bye.
