Motor imagery refers to the psychological realization of movements without movement or muscle activity; it is a research hotspot in neurophysiology, neuroimaging, neurology, and psychology; and it is used as a neurological rehabilitation of brain-computer interface technology. The foundation is widely studied. The EEG signal has a millisecond time resolution, which can facilitate the acquisition of neural signal changes in the body movement process. This article has selected the finger and thumb index finger movements as the task of motion imaging. Under the guidance of image and video, the characteristics of the brain electrical shock patterns during the finger-to-finger movement, thumb and forefinger pairing and unfolding were studied. The brain electrical signals of 15 healthy subjects were collected during the experiment, and the signals were analyzed for source location and information flow network. The result of source location analysis indicated that the activation of brain regions was mainly in the contra lateral SMA region, PMC region, and M1 region. However, the subjects were right-handed, so the characteristics of contra lateral activation were not obvious in the right-handed motor imaging process. Imagine that the opposite side of the pinching process is more powerful than the imaginary finger and the activation range is wider. The results of dynamic information flow of EEG signals at different stages of motor imaging show obvious contra lateral activation characteristics. The DTF analysis of the left-hand motion imaging finger pair pinching and unfolding shows that the information flow mainly flows from the right side of the brain to the left in the imagination process, while the right-hand motion imaging process is opposite, and the information flow mainly flows from the left side of the brain to the right side. The left hand imagines that the finger-to-pinch process is simpler and the number of connections is smaller than the information flow of the imagined finger deployment process. The right hand imagines that the finger-to-kneading process looks more complex than the imagined finger-expanding process, and the number of connections is more. However, its more connections appear mainly in the occipital region of the visual stimulus, the information flow in the frontal and parietal lobes is simple, and the number of connections is reduced. This article explores the relationship among the different movement phases of hand motion imaging and the change of EEG signals. The results show that the brain oscillation modes are similar and different, and the finger deployment process is used as a follow-up of the finger-to-kneading process. The process of knowing should be relatively simple.