Please use this identifier to cite or link to this item:
http://hdl.handle.net/2445/199924
Title: | Development of a pipeline for the study of resting-state fMRI abnormalities in neurological disorders |
Author: | Vallvé Salvadó, Anna |
Director/Tutor: | Sala Llonch, Roser |
Keywords: | Enginyeria biomèdica Neurologia Materials biomèdics Electrònica mèdica Treballs de fi de grau Biomedical engineering Neurology Biomedical materials Medical electronics Bachelor's theses |
Issue Date: | 7-Jun-2023 |
Abstract: | The study of brain functional connectivity abnormalities in neurological disorders is not straightforward. The absence of a standardized and well-defined pipeline and the lack of accepted imaging biomarkers give rise to the need to set certain guidelines and common measures to assess the presence of functional abnormalities in neurological disorders. To provide a solution to the current problem, this project studies the whole-brain network dynamics with resting-state functional magnetic resonance imaging (fMRI) data from 49 patients with Post-COVID-19 neurological syndrome, scanned twice, with a 6-month period between scans. These data are firstly preprocessed to further undergo a node-based (or data-driven) study, more specifically group Independent Component Analysis (ICA). Several decompositions of different dimensionalities are tested to find the optimal range number of independent components according to several levels of granularity (i.e., separation of the networks into subnetworks). The outcome is a set of spatial maps and timecourses, one for each independent component. Then, dual regression is needed to set the group-ICA maps to each individual subject, resulting in a collection of spatial maps and timecourses for each component and each subject. In parallel with dual regression, the independent components must be classified between noise and resting state networks (RSN) and subnetworks. Hierarchical maps are helpful to visualize this classification. Group data comparisons between two time points are carried out to finally identify biomarkers. Four biomarker candidates (i.e. quantitative individual measures obtained from the analyses) are studied: BOLD signal amplitude, full correlation, partial correlation, and covariance between brain regions. In addition, we implement different data-representation approaches that can help to understand the localization of the effects from the subnetwork to the network level. This representation might be helpful to interpret the findings from the point of view of cognitive and mental processes. According to the goal of the project, the resulting pipeline and the extracted biomarkers can be used for analyzing resting-state fMRI data from other neurological disorders. |
Note: | Treballs Finals de Grau d'Enginyeria Biomèdica. Facultat de Medicina i Ciències de la Salut. Universitat de Barcelona. Curs: 2022-2023. Tutor/Director: Sala Llonch, Roser |
URI: | http://hdl.handle.net/2445/199924 |
Appears in Collections: | Treballs Finals de Grau (TFG) - Enginyeria Biomèdica |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
TFG_Vallve_Salvado_Anna.pdf | 19.5 MB | Adobe PDF | View/Open |
This item is licensed under a Creative Commons License