NeuroMark-DyFICA: NeuroMark dynamic frequency-informed ICA with high-frequency spatial filtering for windowed fMRI...

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Psychiatric Disorders

Background:

• Functional brain networks exhibit fine-scale spatial dynamics, often missed by conventional Independent Component Analysis (ICA).
• These subtle variations in connectivity may indicate transient brain function alterations and disordered states.
• Detecting these high-frequency dynamics is crucial for understanding brain function and pathology.

Purpose of the Study:

• To introduce NeuroMark-DyFICA, a novel framework for enhanced detection of spatiotemporal variability in fMRI data.
• To capture transient, fine-scale reconfigurations of brain network topography.
• To identify potential biomarkers for psychiatric disorders by analyzing high-frequency brain dynamics.

Main Methods:

• Developed NeuroMark-DyFICA, integrating dynamic NeuroMark ICA across sliding windows, high-pass spatial filtering, and group-level ICA.
• Estimated time-varying, spatially constrained networks and refined dynamic components.
• Validated the framework using 2D simulations and applied it to resting-state fMRI data from schizophrenia patients and healthy controls.

Main Results:

• NeuroMark-DyFICA successfully detected subtle spatial shifts in simulations that conventional ICA missed.
• Analysis of schizophrenia fMRI data revealed abnormalities in network dynamics, including state imbalances and altered state convergence.
• Six representative brain systems (thalamus, auditory, visual/fusiform, middle frontal, default mode, cerebellum) showed significant alterations.

Conclusions:

• NeuroMark-DyFICA establishes a reproducible latent space for high-frequency brain dynamics.
• The method reveals fine-grained spatiotemporal disruptions in brain networks, offering mechanistic insights into psychiatric disorders.
• This approach holds promise for identifying novel biomarkers for conditions like schizophrenia.