Neurovascular Coupling Non-Invasively Characterized by Simultaneous DC-Magnetoencephalography and Time-Resolved Near-Infrared Spectroscopy

Functional brain imaging methods map neuronal activations indirectly through the accompanying neurovascular response. However, awareness is increasing that the link between neuronal and vascular task-related responses is all but a simple linear transform. Recently, we have shown that in a technically complementary approach neurovascular coupling can be analyzed non-invasively in the human brain in principle by simultaneous DC-magnetoencephalography (DC-MEG with brain-to-sensor modulation) and near-infrared spectroscopy (NIRS). As an extension to the first non-selective NIRS and DC-MEG recordings, we demonstrate here that time-resolved multichannel NIRS in combination with DC-MEG permits us to analyze neurovascular coupling in the human cortex. Simultaneously, DC-MEG signals and time-resolved multichannel NIRS at three wavelengths (687 nm, 803 nm, 826 nm) allowing for depth-selective analysis of absorption changes were recorded over the left primary motor cortex hand area in healthy subjects during finger movements of the right hand (30s followed by 30s rest; n=25 periods). DC fields and NIRS parameters (deoxy-Hb and oxy-Hb) followed closely the motor task cycles revealing statistically significant differences between periods of finger movements and rest. Analysis of variance of distributions of photon times of flight demonstrated changes of hemoglobin concentration originating from a deeper compartment of the head, i.e., the cortex. Notably, the leading slope of the DC-MEG response was found to be steeper and correspondingly the 50% level of its maximum was reached about 1s to 3s earlier compared to the NIRS responses. These results indicate that the vascular response, analyzed here in a simple motor task, is slower compared to the neuronal activation. Taken together, this dual approach provides a new opportunity to analyse non-invasively the “hemodynamic inverse problem“ which refers to the challenge of making valid and precise estimates of underlying cortical neuronal activity from measured hemodynamic responses, e.g., in NIRS and fMRI.