Title: Compressive neural activity detection with fMR images using Graphical Model Inference

Authors: Chuan Li, Qi Hao, Weihong Guo, Fei Hu

Addresses: Department of Electrical and Computer Engineering, The University of Alabama, Tuscaloosa, AL 35487, USA. ' Department of Electrical and Computer Engineering, The University of Alabama, Tuscaloosa, AL 35487, USA. ' Department of Mathematics, Case Western Reserve University, Cleveland, OH 44106, USA. ' Department of Electrical and Computer Engineering, The University of Alabama, Tuscaloosa, AL 35487, USA

Abstract: In this paper, a neural activity detection framework using fMRI data is proposed. The framework consists of three stages: Prediction: Predicting regions of interest associated with an extraordinary amount of neural activities through Temporal Clustering Analysis (TCA). Modelling: Categorising fMRI signals related to neural activities into event prototypes through Linear Predictive Coding (LPC). Inference: Determining the types of neural activities in terms of activation, deactivation and normality through graphical model based Bayesian inference. The experiment results demonstrate the advantages of this approach in terms of computational cost and robustness against artefacts.

Keywords: neural activity detection; fMR images; graphical modelling; LPC; linear predictive coding; VBGMM; variational Bayesian Gaussian mixture models; functional magnetic resonance imaging; functional MRI; temporal clustering analysis; TCA.

DOI: 10.1504/IJCBDD.2010.038024

International Journal of Computational Biology and Drug Design, 2010 Vol.3 No.3, pp.187 - 200

Published online: 11 Jan 2011 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article