The neurovascular coupling bears properties of a feedforward and feedback regulative mechanism.

Temporal profiles of mediators involved in the neurovascular coupling bear feedforward and feedback characteristics, which are supported by mathematical modeling of evoked hemodynamic responses. However, cerebral autoregulation was expressed as a feedback control system. Therefore, question of overdetermination of the neurovascular coupling model arises. Addressing this issue, we analyzed both models for their appropriateness in describing the neurovascular coupling. Visually evoked flow velocities were recorded from healthy volunteers (aged 24.7 +/- 1.6 SD; 9 females, 11 males) with transcranial Doppler in the posterior cerebral artery. Control system parameters of the two models were specified according to the least-square-fitting technique. Mean square errors between measured and modeled curves and Akaike's information criterion (AIC) supported the feedforward-feedback model. Mean square differences decreased from 27.2 +/- 37.9 to 2.3 +/- 2 and the AIC from 2.7 +/- 0.6 %(2) to 2 +/- 0.3 %(2). The feedforward element increased the accuracy of the control system model in describing the fast initial response. Biologically, the parameter decreases the initial mismatch between the fast neuronal but slow vascular response because of visual activation.