Back to search

Boston University

Time-varying functional connectivity and dynamic neurofeedback with MEG: methods and applications to visual perception

Abstract

dc:description.abstract

Cognitive function involves the interplay of functionally-separate regions of the human brain. Of critical importance to neuroscience research is to accurately measure the activity and communication between these regions. The MEG imaging modality is well-suited to capturing functional cortical communication due to its high temporal resolution, on the millisecond scale. However, localizing the sources of cortical activity from the sensor measurements is an ill-posed problem, where different solutions trade-off between spatial accuracy, correcting for linear mixing of cortical signals, and computation time. Linear mixing, in particular, affects the reliability of many connectivity measures. We present a MATLAB-based pipeline that we developed to correct for linear mixing and compute time-varying connectivity (phase synchrony, Granger Causality) between cortically-defined regions interfacing with established toolboxes for MEG data processing (Minimum Norm Estimation Toolbox, Brainstorm, Fieldtrip). In Chapter 1, we present a new method for localizing cortical activation while controlling cross-talk on the cortex. In Chapter 2, we apply a nonparametric statistical test for measuring phase locking in the presence of cross-talk. Chapters 3 and 4 describe the application of the pipeline to MEG data collected from subjects performing a visual object motion detection task. Chapter 5 focuses on real-time MEG (rt-MEG) neurofeedback which is the real-time measurement of brain activity and its self-regulation through feedback. Typically neurofeedback modulates directly brain activation for the purpose of training sensory, motor, emotional or cognitive functions. Direct measures, however, are not suited to training dynamic measures of brain activity, such as the speed of switching between tasks, for example. We developed a novel rt-MEG neurofeedback method called state-based neurofeedback, where brain activity states related to subject behavior are decoded in real-time from the MEG sensor measurements. The timing related to maintaining or transitioning between decoded states is then presented as feedback to the subject. In a group of healthy subjects we applied the state-based neurofeedback method for training the time required for switching spatial attention from one side of the visual field to the other (e.g. left side to right side) following a brief presentation of a visual cue. In Chapter 6, we used our pipeline to investigate training-related changes in cortical activation and network connectivity in each subject. Our results suggested that the rt-MEG neurofeedback training resulted in strengthened beta-band connectivity prior to the switch of spatial attention, and strengthened gamma-band connectivity during the switch. There were two goals of this dissertation: First was the development of the MATLAB-based pipeline for computing time-evolving functional connectivity analysis in MEG and its application to visual motion perception. The second goal was the development of a real-time MEG neurofeedback method to train the dynamics of brain states and its application to a group of healthy subjects.

Degree

thesis:*
Name thesis:degree_name
Doctor of Philosophy
Level thesis:degree_level
doctoral
Discipline thesis:degree_discipline
Biomedical Engineering
Grantor
Boston University
Year dc:date.issued
2017

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Rana, Kunjan Dinesh

Subjects

dc:subject × 1

Rights

Language dc:language.iso
en_US

Identifiers

dc:identifier.*
Handle dc:identifier.uri
https://hdl.handle.net/2144/27050

Chain of custody

source
Harvested from
Boston University
Base URL
open.bu.edu/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
related terms
citation

Rana, Kunjan Dinesh. Time-varying functional connectivity and dynamic neurofeedback with MEG: methods and applications to visual perception. doctoral thesis, Boston University, 2017. https://hdl.handle.net/2144/27050