A Novel Network Connectivity Measure with Application to Multimodal Brain Imaging Study 公开
Yang, Xinyi (Spring 2019)
Abstract
Background: Network-oriented analysis on functional connectivity from functional magnetic resonance imaging (fMRI) data is not easy due to the large scale of fMRI data. The traditionally used Pearson correlation is not ideal because it is sensitive to indirect connectivity. The recently proposed partial correlation also has shortcomings because of both the loss of the "indirect" connectivity information and the requirement of a prescribed tuning parameter. Thus, a novel measure for functional network connectivity is proposed, in order to better discribe functional connevtivity and get rid of the predetermined tuning parameter.
Methods and Analysis: The distance derived from CLIME is used to construct the global structure in the brain. To overcome the problem of tuning parameter selection, we integrate out the effect of the tuning parameter. The intraclass correlation coefficient is calculated on Kirby21 data. Also, a multimodal brain imaging analysis is conducted based on the new functional connectivity measure for fMRI and DTI data from the PNC study with 515 subjects.
Results and Conclusion: We found that the novel measure of functional connectivity has a stronger relationship with structural connectivity than conventional Pearson correlation functional connectivity, which suggests our approach more effectively reveals neurologically related activity in functional magnetic resonance imaging. The intraclass correlation coefficient shows that the robustness of this new measure are at a similar level compared to the existing measures Pearson correlation and partial correlation. Thus, we see that the novel functional connectivity measure not only shares similarity with existing popular functional connectivity measures, but also performs better in terms of representativeness in multimodal analysis.
Table of Contents
1 Introduction.................................... 1
2 Methods.................................... 2
2.1 Notation.................................... 2
2.2 Pathway-embedded functional connectivity measure . . . . . . . . . . . . . 3
2.3 Algorithm................................... 4
3 Analysis ....................................5
3.1 The Choice of Transformation Function ................... 5
3.2 Reproducibility ................................ 5
3.3 Multimodality................................. 6
4 Results.................................... 6
4.1 Visualization ................................. 6
4.2 The Choice of Transformation Function ................... 7
4.3 Reproducibility ................................ 7
4.4 Multimodality................................. 8
5 Discussion.................................... 9
About this Master's Thesis
| School | |
|---|---|
| Department | |
| Subfield / Discipline | |
| Degree | |
| Submission | |
| Language |
|
| Research Field | |
| 关键词 | |
| Committee Chair / Thesis Advisor | |
| Committee Members |
Primary PDF
| Thumbnail | Title | Date Uploaded | Actions |
|---|---|---|---|
|
|
A Novel Network Connectivity Measure with Application to Multimodal Brain Imaging Study () | 2019-04-09 15:15:55 -0400 |
|
Supplemental Files
| Thumbnail | Title | Date Uploaded | Actions |
|---|