In 1980s, the brain imaging technology used to analyze neural activity indexes had a great influence on neuroscience. At present, the two most common applications in cognitive neuroscience research are positron emission tomography (PET) to measure brain energy consumption, and functional magnetic resonance imaging (fmri) to measure cerebral blood flow oxygen consumption. The above two should be used to observe the activity patterns of various cortical areas when the human brain performs advanced functions. Other common applications include electroencephalography (EEG), magnetoencephalography (MEG) and near infrared spectroscopy. The nerve activity took about 6 seconds, and the curve of BOLD slowly reached its peak. But the change of BOLD signal is so lagging behind, how can it quickly reflect the ever-changing world and psychological changes? This problem deeply puzzles many fMRI researchers. So the algorithm in this paper solves this problem.