Exploring functional brain dynamics based on Fisher linear discriminant criterion

Previous studies have demonstrated that the brain functional connectivity undergoes striking temporal dynamics. Modeling dynamically the brain functional connectivity has become not only an urgent and important work, but also a new direction for exploring brain functional research. For this motivation, a novel method for exploring functional brain dynamics based on Fisher linear discriminant (FLD) criterion is proposed, which is not required any prior knowledge. The proposed method has been evaluated and validated by a real block-design task-based functional Magnetic Resonance Imaging (fMRI) datasets of 19 subjects, and the results implied our method had a good performance to detect change points.