Frequency Domain Analysis of Cerebral Blood Flow Velocity and its Correlation with Arterial Blood Pressure

Abstract

We applied frequency domain analysis to detect and quantify spontaneous fluctuations in the blood flow velocity of the middle cerebral artery (MCAFV). Instantaneous MCAFV of normal volunteers was detected using transcranial Doppler sonography. Spectral and transfer function analyses of MCAFV and arterial blood pressure (ABP) were performed by fast Fourier transform. We found the fluctuations in MCAFV, like ABP, could be diffracted into three components at specific frequency ranges, designated as high-frequency (HF, 0.15 to 0.4 Hz), low-frequency (LF, 0.04 to 0.15 Hz), and very low-frequency (VLF, 0.016 to 0.04 Hz) components. The HF and LF components of MCAFV exhibited high coherence with those of ABP, indicating great similarity of MCAFV and ABP fluctuations within the two frequency ranges. However, it was not the case for the VLF component. Transfer function analysis revealed that the ABP-MCAFV phase angle was frequency-dependent in the LF range (r = -0.79, P < 0.001) but not in the HF range. The time delay between LF fluctuations of ABP and those of MCAFV was evaluated as 2.1 seconds. We conclude that in addition to traditional B-wave equivalents, there are at least two different mechanisms for MCAFV fluctuations: the HF and LF fluctuations of MCAFV are basically secondary to those of ABP, and cerebral autoregulation may operate efficiently in LF rather than HF range. Frequency domain analysis offers an opportunity to explore the nature and underlying mechanism of dynamic regulation in cerebral circulation.

Spectral Analysis of Event-Related Hemodynamic Responses in Functional Near Infrared Spectroscopy

“…Our analysis has covered up to 330 mHz range since this was the cut-off frequency of the built-in RC-filter in the fNIRS device. These bands are found to be very similar to those in related studies (Hu et al, 1999;Kuo et al, 1998;Obrig et al, 2000a;Franceschini et al, 2000;Toronov et al, 2000).…”

“…In several other studies, three main frequency bands of interest have been identified for cerebral hemodynamics: a very low frequency VLF (8-33 mHz), a low frequency LF (around 100 mHz) named as the Mayer waves or V-signal (Mayhew et al, 1999;Obrig et al, 2000a), and a high frequency component HF (around 250 mHz), the latter being definitely synchronous with breathing rate (Kuo et al, 1998). Similarly, we conjecture that each of the canonical bands is associated with one or more of the physiological activities related to hemoglobin concentrations.…”

“…The lowest frequency A band (0-0.03 Hz) is mainly responsible for the slow baseline signal that is thought to be reflecting the very slow vasomotor activity due to heart rate fluctuations and thermoregulation Francheschini et al, 2000;Giller et al, 1999). In fact, reports on the frequency content of such fluctuations have identified this signal as being the phasic dilation and contractions of "the small regulating arteries, and these vasomotor waves produce fluctuations in cerebral blood volume, which are eventually reflected in the intracranial pressure" (Kuo et al, 1998). Similarly, in many fMRI studies, the Gamma function model has been used to parameterize the brain hemodynamic response.…”

“…Since the B-band can at most accommodate one such spectral line, we merge the two bands B and C, to cover the 30-330 mHz band range. Previous researchers (Kuo et al, 1998;Giller et al, 1999;Hu et al, 1999) have also excluded the A-band since it was mostly mirroring the fNIRS baseline activity. Another plausible argument for excluding the A-band is that it is inherently a nonstationary process, which obviates signalprocessing tools requiring stationarity.…”

“…Several researchers used spectral estimation techniques and transfer function analysis to relate the brain hemodynamic response to various stimuli and physiological events (Villringer and Chance, 1997;Svensen et al, 2000;Wobst et al, 2001;Hu et al, 1999;Kuo et al, 1998;Giller et al, 1999). Most studies that are performed by transcranial Doppler sonography suggest low frequency domination due to cerebral autoregulation of hemodynamical activity where the increase in acidity levels (increase on carbondioxide content) leads to this very slow vasomotor activity forming the Mayer waves or the V-signal (Obrig et al, 2000a).…”