Modified Principal Component Analysis in sliding windowed fMRI data

Khairi, N.M; Wilkes, D.M.; Ding, Zhaohua

doi:10.1109/southeastcon42311.2019.9020664

Cited by 1 publication

(1 citation statement)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The following conditional probabilistic group analysis then generated a sequence of brain maps describing the whole brain's avalanche evolution. The second method used the modified Principal Component Analysis for Sliding Window (mPCASW) 9 to decompose the signal into components and extract the sequence within components that have positive extremal curve patterns around the avalanches. We then exploited the sequential pattern behavior using the selected components.…”

Section: Introductionmentioning

confidence: 99%

Exploring timing patterns around the phase transition of the brain system using resting-state fMRI

Guo

Khairi

Wilkes

2023

Medical Imaging 2023: Image Processing

View full text Add to dashboard Cite

The brain is a complex system with the characteristic of transitioning between the ordered and disordered phases around an optimal point, criticality. Many studies focus on exploring the brain behaviors around criticality. However, complexity and variability in the criticality also make the work more challenging. Here, we are intrigued to investigate the highly organized phase, the avalanche or avalanche peaks, which can also be represented as the few gigantic peaks in the global fMRI signal. We explored different timing patterns around the rapid phase transition of the avalanche peaks using statistical and image processing methods. The first approach measures the relative peaking time to the avalanche peak throughout the brain. The following conditional probabilistic group analysis generates a sequence of brain maps describing avalanche evolution. The second method used the modified Principal Component Analysis for Sliding Window (mPCASW) to decompose the signal into components and extract the ordering of components showing positive quadratic patterns around the avalanche. The conditional probabilistic results from the first approach identified representative regions at each stage and showed different trajectories in separate regions. We also found that, in most of the primary components of the mPCASW method that behaved in specific patterns around the avalanche, they revealed the ordering in the Brodmann area regions' activations. These pattern behaviors are consistent across the population. We believe there could be a meaningful interpretation that can be extracted in the future.

show abstract

Section: Introductionmentioning

confidence: 99%