Short Introduction to Bayesian Inference for Neuroimaging

by Michael Chappell, Mark Woolrich, Mark Jenkinson

It is not necessary to understand all of the details of the algorithms used to analyse neuroimaging data. However, some understanding of the concepts that lie behind these algorithms can be helpful to be able to design good experiments, make appropriate analysis choices, and interpret results carefully. In this Short Introduction we will outline the basics of Bayesian Inference, a popular mathematical framework for data analysis and machine learning that is used in many areas of neuroimaging analysis. This text is one of a number of appendices to the Oxford Neuroimaging Primers, designed to provide extra details and information that someone reading one of the primers might find helpful, but where it is not crucial to the understanding of the main material. This appendix specifically addresses the principles that underpin Bayesian Inference, as it is used in neuroimaging. In it we seek to go into more detail than we might in one of the primers, for those who want to understand more about how Bayesian Inference can be used for data analysis. In turn, this appendix also provides a high level introduction to individuals who are interested in developing their own Bayesian Inference methods, or find they need to select between different methods in a specific application.

43 pages, Kindle Edition

Published February 24, 2020