On Hippocampus Associative Modeling by Approximating Nonlinear Kullback-Leibler Sparsity Constraint

Identifying dementia at early phase of mild cognitive impairment (MCI) is essential for diagnosis and intervention for Alzheimer's disease (AD). Automated diagnosis of resting-state functional magnetic resonance imaging (rs-fMRI) leads to early identification of dementia. Early detection of dementia at the stage of mild cognitive impairment (MCI) is limited to effective spatial-temporal dependency. The retrieval of functional connectivity among various hubs of human brain estimates the neuronal health and disease progression. Due to motion related artifacts at acquisition time, noise intervention and for many other reasons the sparse constraint becomes predominant for Rf-MRI. In this paper a unique Kullback-Leibler (K-L) divergence based sparse constrained regression model is proposed. The proposed model creates a framework that can identify and analyze connectivity between hippocampus and other significant region of interests (ROI) of brain. The results show a promising improvement in connectivity measurements of hippocampus for resting state functional MRI. The outcome of simulated results also appears in form of correlation matrix which shows great efficiency of dealing connectivity constraints. The entire process leads to trace neurodegenerative disease related malfunctions, especially MCI more precisely.