Comparison of Classification and Dimensionality Reduction Methods Used in fMRI Decoding

In the last few years there has been growinginterest in the use of functional Magnetic Resonance Imaging(fMRI) for brain mapping. To decode brain patterns in fMRIdata, we need reliable and accurate classifiers. Towards this goal,we compared performance of eleven popular pattern recognitionmethods. Before performing pattern recognition, applying thedimensionality reduction methods can improve the classificationperformance; therefore, seven methods in region of interest(ROI) have been compared to answer the following question:which dimensionality reduction procedure performs best? Inboth tasks, in addition to measuring prediction accuracy, weestimated standard deviation of accuracies to realize morereliable methods. According to all results, we suggest usingsupport vector machines with linear kernel (C-SVM and ν-SVM),or random forest classifier on low dimensional subsets, which isprepared by Active or maxDis feature selection method toclassify brain activity patterns more efficiently.

Keywords:

Brain Image analysis , Functional MRI , Classification , Dimensionality Reduction