Functional MRI (fMRI) and near infrared spectroscopy (NIRS) are two non-invasive imaging modalities used to study human brain function. However, they can only measure neuronal activation indirectly through neurovascular coupling. Other non-neuronal physiological fluctuations can also influence their signals. In this talk, I will introduce a novel method to identify a physiological low frequency oscillation (LFO) in fMRI signal through concurrent fMRI/NIRS study. It is demonstrated that this LFO is a dynamic signal, which travels with the blood throughout the brain. Moreover, we demonstrated that the same LFO was also found in the peripheries through another concurrent fMRI/NIRS experiment. This finding showed the systemic feature of this wave and indicated its potential extracerebral origin. As application, we demonstrated the method to convert this “confounding” systemic signal into useful perfusion biomarker. Using this LFO and its temporal shifts, dynamic cerebral blood flow map can be obtained from regular resting state fMRI scans, which allows simultaneous assessment of functional networks and perfusion. Furthermore, novel methods based on the similar procedure were developed to: 1) track cardiac pulsation wave in the fMRI data; 2) to denoise fMRI data and 3) to assess cerebral blood flow recovery from Moyamoya patients after surgery.