Identification of MCI using optimal sparse MAR modeled effective connectivity networks.

摘要: Capability of detecting causal or effective connectivity from resting-state functional magnetic resonance imaging (R-fMRI) is highly desirable for better understanding the cooperative nature brain. Effective provides specific dynamic temporal information R-fMRI time series and reflects directional influence one brain region over another. These influences among regions are normally extracted based on concept Granger causality. Conventionally, inferred using multivariate autoregressive (MAR) modeling with default model order q = 1, considering low frequency fluctuation series. This assumption, although reduces complexity, does not guarantee best fitting at different regions. Instead order, we propose to estimate optimal upon MAR distribution characterize these each region. Due sparse networks, an orthogonal least square (OLS) regression algorithm incorporated minimize spurious connectivity. networks proposed applied Mild Cognitive Impairment (MCI) identification obtained promising results, demonstrating importance relationships between neurodegeneration disorder identification.