Comparison of LFOs among Different Hemodynamic Parameters under Different Physiological Conditions Among these hemodynamic parameters, LFOs of rCBF, Δ[HbO2] and ΔTHC had relatively stable phase shifts (std(Φ) ≤ 88.9°) and were detected at reasonable success rates (≥ 67%) under the three physiological conditions. Compared to other variables, LFOs of Δ[Hb] were much weaker, which agreed with previous findings [7, 148]. As a consequence, the phase shifts of Δ[Hb] had large variations. The weak LFOs of Δ[Hb] were likely caused by the low amplitude of Δ[Hb] in cerebral blood. Cerebral blood has a high oxygen saturation varying from almost 100% in arterial blood to ~60% in venous blood, indicating the small percentage of [Hb] in cerebral blood [154]. It is also possible that the cerebral venula (containing more [Hb]) is less responsible to blood pressure variations than the arteriole (containing more [HbO2]) [155]. As a result, Δ[Hb] exhibited the least magnitude and SNR of LFOs, and was therefore excluded from the following discussions. LFOs at rest were weak, resulting in low SNRs [150]. As a result, success rates were low and LFO frequencies, phase shifts, and phase differences between the two hemispheres had large variations (SDs) at rest. HUT and paced breathing significantly enhanced LFOs through different mechanisms. During paced breathing, LFOs at 0.1 Hz were elevated by respiration-induced MAP oscillation [150]. However, there were large inter-subject variations in phase shifts of cerebral hemodynamics, which might be attributed to the heterogeneous responses to paced breathing among subjects. The hypocapnia might be induced by relatively deep and slow breathing pattern used in the present study, resulting in increased variations in cerebral hemodynamics [150]. LFOs during HUT were enhanced probably through the increased sympathetic nervous activity [142, 151]. Previous study has demonstrated that a single burst in sympathetic nervous system can initiate a cycle of increasing and decreasing arterial pressure at ~0.1 Hz via the baroreflex feedback loop [156]. The sympathetic nervous activity bursts in series each ~10 s during HUT [157], which can significantly enhance LFOs of MAP through baroreflex resonance. Moreover, compared to the paced breathing protocol, the HUT protocol is easier to control and more objective. Overall, signals obtained during HUT were most stable (smallest standard deviations) and robust (highest success rates) among the three physiological conditio...
