Functional MRI Time Series Generation via Wavelet-Based Image Transform and Spectral Flow Matching for Brain Disorder Identification
Researchers have developed a new framework called Dual-Spectral Flow Matching (DSFM) to generate functional MRI (fMRI) time series data. This method addresses limitations in current generative models by better replicating the non-stationary and dynamic nature of raw BOLD signals. DSFM utilizes wavelet and cosine transforms to capture multi-scale variations and energy compaction, ultimately improving downstream brain disorder classification tasks. AI
IMPACT Enables more robust training of AI models for brain disorder identification by providing synthetic fMRI data.