This project aimed to further my intuitive understanding of fMRI data. Around 20 interactive/static figures of various statistics of raw fMRI data, confounds and atlased data were produced. Special efforts have been made to make the analysis highly and easily reproducible.
This project aimed to develop a BIDS App for performing Connectome-Wide Association Studies (CWAS) on fMRI connectivity matrices. The result is a GitHub repository that can be installed via pip, enabling analyses on BIDS-formatted connectomes. Integration tests ensure the pipeline runs reliably, and a dedicated website provides full documentation and example outputs.
This project explores how different fMRI analysis methods reveal distinct aspects of the neural representation of fear, including a mass univariate approach (GLM), a machine learning (decoding) approach, and representational similarity analysis (RSA) approach. While GLM identified some expected activation patterns and machine learning failed to decode fear ratings reliably, RSA revealed modest but significant structure in frontal regions, highlighting the value of methodological triangulation in cognitive neuroscience.
This study examines dorsolateral prefrontal cortex (dlPFC) and posterior parietal cortex (PPC) connectivity and dlPFC BOLD time series in ADHD versus typically developing (TD) children during the cued stop-signal task (CSST) using fMRI data from OpenNeuro ds005899. It is hypothesised that stronger dlPFC-PPC connectivity will be found in the ADHD group.
This mini-project for the 2025 Brainhack School is part of my PhD dissertation on Late-Life Cognitive Heterogeneity, where I examine the neural correlates of cognitive dispersion – a measure of within-individual variability – using neuropsychological and fMRI data from the Midnight Scan Club dataset (OpenNeuro ds000224).
This project aimed to explore tools and techniques used to analyze fMRI brain activation at the first level using data from an open-access dataset. We produced a comprehensive Jupyter Notebook that provides a step-by-step guide to running the analyses, including applying the GLM, defining contrasts, and generating various brain activation maps.
This project aims to examine how the brain supports two types of inference—active and passive—during the process of posterior belief integration. I applied trial-by-trial analysis to track participants’ learning trajectories over time.
Can we predict music genres based on fMRI connectivity patterns alone? This project explores a single-subject decoding approach using ROI-to-ROI correlation matrices and machine learning classifiers on OpenNeuro dataset ds003720.
This project applies machine learning to decode perceived emotions from fMRI data using ROI-based features. Data from the ds003548 OpenNeuro dataset are analyzed, with task labels extracted from events files. ROI time series are extracted using the MIST 64-ROI atlas, and mean signals during emotion blocks are classified using linear SVM. The goal is to distinguish between six conditions (happy, sad, angry, neutral, blank, scrambled), demonstrating key concepts and challenges in neuroimaging-based classification.
The project focused on extracting activities from functional images in a previous study about neural representation of melody-transposition. Using parcellated brain atlas as a mask, the BOLD signals underwent univariate analysis to look for effects in error detection or music-like stimulus-related brain regions/
Copyright (c) 2025, BrainHack School; all rights reserved.
Template by Bootstrapious. Ported to Hugo by DevCows.