Orhan Firat, Itir Onal, Emre Aksan, Burak Velioglu, Ilke Oztekin, Fatos T. Yarman Vural
Data mining and Machine Learning, Magnetic ResonanceImaging, Medical Image Processing, Brain State Decoding,Functional Connectivity, MVPA.
Functional Magnetic Resonance Imaging (fMRI) data con- sists of time series for each voxel recorded during a cog- nitive task. In order to extract useful information from this noisy and redundant data, techniques are proposed to select the voxels that are relevant to the underlying cog- nitive task. We propose a simple and efficient algorithm for decoding the brain states by modelling the correlation patterns between the voxel time series. For each stimulus during the experiment, a separate functional connectivity matrix is computed in voxel level. The elements in con- nectivity matrices are then filtered out by making use of a minimum spanning tree formed using a global connectivity matrix for the entire experiment in order to reduce dimen- sionality. For a recognition memory experiment with nine subjects, functional connectivity matrices are computed for encoding and retrieval phases. The class labels of the re- trieval samples are predicted within a k-nearest neighbour space constructed by the traversed entries in the functional connectivity matrices for encoding samples. The proposed method is also adapted to large scale functional connectiv- ity tasks by making use of graphics boards. Classification performance in ten categories is comparable and even bet- ter compared to both classical and enhanced methods of multi-voxel pattern analysis techniques.
Important Links:
Go Back
