Different strategies to initiate and maintain hyperventilation: their effect on continuous estimates of dynamic cerebral autoregulation