ABSTRACT
Contents Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246 13.1 Autism and the importance of brain connectivity . . . . . . . . . . 246 13.2 Review of ndings in brain connectivity and autism . . . . . . . . 249
13.2.1 Structural connectivity studies . . . . . . . . . . . . . . . . . . . 249 13.2.2 Functional and effective connectivity studies . . . . . . . 251
13.2.2.1 Resting-state studies. . . . . . . . . . . . . . . . . . . . 252 13.2.2.2 Task-based studies.. . . . . . . . . . . . . . . . . . . . . 253
13.3 How to characterize brain networks. . . . . . . . . . . . . . . . . . . . . 255 13.4 Connectivity explanation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258 13.5 Measuring functional or effective connectivity with
graph theory methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259 13.6 Measuring effective connectivity with Granger causality . . . . 263
13.6.1 Denition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264 13.6.2 Mathematical form of GC . . . . . . . . . . . . . . . . . . . . . . 265 13.6.3 Conditional Granger causality . . . . . . . . . . . . . . . . . . . 265 13.6.4 Granger causality limitations . . . . . . . . . . . . . . . . . . . . 266
13.7 Imaging GC connectivity in cases with autism spectrum disorder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266 13.7.1 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
13.7.1.1 Case 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 13.7.1.2 Case 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270 13.7.1.3 Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
13.8 Clinical applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273 13.8.1 Treatment case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
13.9 Future directions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Abstract Autism spectrum disorder (ASD) has a strongly neurobiological component. While the search for single brain structures to explain the symptoms of ASD has been ineffective, research over the past decade has shown that there are strong indicators for neural connectivity anomalies that do relate to the challenges that these persons face. Information is presented regarding types of connectivity with effective connectivity being the goal in the assessment of ASD-related difculties. Using a graph theory model, we explore aspects of connectivity and its assessment related to ASD. The measure of effective connectivity with EEG technology along with the use of Granger causality is presented as a means of producing connectivity estimates that are auto-regressive, predictive and demonstrate causal and reciprocal inuences across multiple brain regions. Several case studies are presented with their graphical ndings to demonstrate the use and utility of this approach in the understanding of brain mechanisms impacting ASD.
