Ivor Cribben

This study compares static, sliding window, and change point dynamic functional connectivity methods for classifying early mild cognitive impairment using resting-state fMRI, finding cpDFC outperforms others with 91.17% accuracy.

This paper introduces a Vine Copula Change Point model to estimate dynamic functional connectivity in fMRI data, showing it can capture non-Gaussian and non-linear dependencies between brain regions.

This paper introduces CCID, a change-point detection method for dynamic functional connectivity, and shows it performs well on simulated and fMRI data.

This study examines how healthy cognitive aging affects resting-state fMRI signal measures, finding age-related changes in network amplitude and inter-network connectivity but some architectural invariance.

This paper introduces a nonparametric anomaly detection framework for dynamic networks, demonstrating improved accuracy on simulated and real-world data including fMRI and communication graphs.

This paper introduces a longitudinal functional connectivity network model for fMRI data, accounting for autocorrelation and subject variability, and applies it to Alzheimer's Disease Neuroimaging Initiative data.

This paper evaluates sparse graphical models for functional brain networks, finding SCAD with BIC and CV is best, and autocorrelation harms estimation but is mitigated by CV.