Structural Brain Variability Over Time – What is Driving it? :An Exploratory Longitudinal MRI Study of Extensive Sampling

Keywords

Loading...
Thumbnail Image

Issue Date

2021-08-25

Language

en

Document type

Journal Title

Journal ISSN

Volume Title

Publisher

Title

ISSN

Volume

Issue

Startpage

Endpage

DOI

Abstract

This study used “The Quantified Self Dataset” where a single subject was sampled for 93 weeks with weekly MR scans and a large range of daily markers to exploratorily investigate how brain structure of a healthy brain varies over this timespan. Conventional T1w structural imaging as well as quantitative structural imaging sequences were used to investigate how changes in biomarkers like sleep dynamics, blood pressure, weight, body temperature, blood oxygen saturation and aging affects these structural changes. Additionally, differences between the three structural imaging sequences (MPRAGE, MP2RAGE and MP2RAGE-Sparse) were evaluated to see which sequence showed the most sensitivity to these changes and which pipeline was more appropriate for a given sequence dataset. A GLM was performed on VBM, SBM and raw intensity maps as well as an ICA was performed. GLM only found an effect of time. ICA decomposed structural changes components with in meaningful timecourses and spatial maps, showing prominent changes surrounding the ventricles, cardiac changes, white matter intensity changes and changes in tissue borders. Changes in tissue volume as well as tissue composition were found where GM decreased in volume and WM appears to be more myelinated. Results also showed an increase in ventricular volume where the ventricles expanded. Comparing results of different structural MR-sequences, MPRAGE seems best suited for GLM on VBM and SBM measurements and MP2RAGE seems best suited for the GLM on raw intensity images. Additionally, MP2RAGE(-Sparse) seems best suited for ICA. Extensive sampling over the span of a long time was possible because only one subject was used. However, as regressor quality seems to have been lacking in this study, improvement of the quality of sampling seems a priority.

Description

Citation

Supervisor

Faculty

Faculteit der Sociale Wetenschappen