Spatial changes in local activation in patients on the autism spectrum using resting state BOLD fMRI and T1w images

Abstract

Background: The pursuit for determining and analyzing circumscribed brain regions, or parcels, has been of major interest for neuroscientists since the late 19th century with increasingly prevalent research that was carried out in the field, particularly by Brodmann and his advances in cytoarchitecture. The recently proposed "Vogt-Bailey Index" was introduced as a metric of the smoothness of cortical feature boundaries in the human brain. Objective: In this project, we aim to quantify the changes in local functional boundaries within the brains of young patients having Autism Spectrum Disorder and compare them to healthy control subjects using this metric. This provides a means to explore the differences in brain functional architecture and can lead to new understanding of the cortical structure of Autism Spectrum Disorder. Research Design: A comparative quantitative research approach was adopted for this project in order to statistically compare the variability between these two groups using the Aspin-Welch test with the Permutation Analysis for Linear Models tool. The results are then contrasted to a similar existing technique that studies local functional connectivity in the human brain called Regional Homogeneity. Results: The results showed a higher brain activity for patients in the regions responsible for effective processing and executive function, planning and sensitivity to sequential action. They showed lower brain activity in regions responsible for visual, auditory, language, sensory, motor signal, parasympathetic and sympathetic processing. Conclusions and Recommendations: The application of the tool was therefore seen as successful in identifying higher- and lower-activity regions of the brain that are aligned with the common symptoms in Autism Spectrum Disorder patients when compared to healthy control subjects. For future research, we recommend the use local data from local health institutions for a more accurate representation of our national patients, together with using different covariates of no interest, increasing the data size and possibly examining other psychiatric disorders of the nervous system. We also recommend amending the tools to output numerical results so that a statistical comparison between tools can be made. For professional practice, we recommend improving the methods to reduce patient movement and motion artifacts while also reduce scanning time without compromising heavily on image quality.