Peripheral vascular function and dynamics of muscle metabolism may be better understood by measuring the temporal relationship of blood flow and oxygen saturation in micro and macrovasculature^1-7. Concurrent acquisition of perfusion, SvO₂, T₂*, and arterial blood flow would offer a comprehensive functional assessment of the peripheral vasculature, may allow for assessment of the specific contribution of macrovascular transit delays (versus primary impairment in microvascular reactivity), and would permit calculation of oxygen consumption ($$$\dot{V}O_{2}$$$) via the Fick principle: $$$\dot{V}O_{2}=CaO_2∙flow∙(SaO_2-SvO_2 )$$$, where CaO₂ is the arterial oxygen content. Thus, the purpose of this project was to develop a technique to dynamically and simultaneously measure flow in the macro and microvasculature, as well as oxygen saturation.Simultaneous measurement of perfusion, arterial blood flow, SvO₂, and T₂*, was achieved with a 3-slice interleaved sequence termed Velocity and Perfusion, Intravascular Venous Oxygen saturation and T₂* (vPIVOT) (Figure 1). The traditional PIVOT sequence comprises a dual-slice acquisition scheme – in which a multi-echo GRE sequence acquires data at a slice downstream from the perfusion slice during the PASL post-labeling delay (PLD)⁴. Here, the framework of PIVOT was expanded to include an upstream velocity-encoded GRE sequence following the PASL EPI acquisition. Additional pulse sequence parameters are: PASL: Slice-selective or non-selective adiabatic inversion with PLD=942 ms, partial Fourier GRE-EPI, acquisition matrix=80×50 (reconstructed to 80×80), FOV=25×25 cm, slice thickness=10 mm, location=isocenter, TR/TE=2000/9 ms; Inferior location multi-echo GRE: acquisition matrix=96×24, FOV=96×96 mm, slice thickness=10 mm, slice location=3 cm distal, TR/TE₁/TE₂/TE₃/TE₄/TE₅=38.1/3.8/7.0/12.3/19.3/26.3 ms, a=15°; Superior location velocity-encoded GRE: acquisition matrix=96×24, FOV=96×96 mm, slice thickness=5 mm, slice location=3-4 cm proximal, TR/TE₁/TE₂=20.0/6.0/9.7 ms, a=10°, VENC=120 cm/s. Both the superior and inferior GRE data were reconstructed to 96×96 via keyhole reconstruction⁸. All parameters are quantified at 4-second temporal resolution. Imaging experiments were performed at 3T with an 8-ch Tx/Rx knee coil. To assess whether the sequence is unbiased in its measurement of each parameter, vPIVOT-derived perfusion, SvO₂ and T₂*, or SvO₂ and flow were compared to those obtained with PIVOT, or a standard velocity-encoded multi-echo GRE (termed OxFlow), respectively. In 5 healthy subjects (4 male, 32.4±6.4 years old) who had previously provided consent, data were acquired with vPIVOT, PIVOT, and OxFlow throughout separate ischemia-reperfusion paradigms, each with 2 min baseline, 3 min occlusion, 4 min recovery. vPIVOT data were then acquired continuously throughout 2 minutes baseline, 2 minutes of resisted dynamic plantar flexion contractions (6 watts) using a custom-built MRI-compatible ergometer, and 8 minutes recovery. Muscle volume was quantified from a 3D SSFP-echo acquisition.Respective quantification methods were used to compute popliteal artery flow, gastrocnemius perfusion⁹, popliteal vein and posterior tibial vein SvO₂¹⁰, and soleus muscle T₂*, and ischemia-reperfusion time courses were parameterized to provide response timing and magnitude⁴. $$$\dot{V}O_{2}$$$ was quantified for the exercise scan using either arterial flow (normalized to muscle mass) or perfusion. For both, SaO₂ was assumed to be 100%, and average SvO₂ was calculated from superior and inferior GRE acquisitions. In Figure 2 the average ischemia-reperfusion time courses for blood flow, perfusion, SvO₂, and T₂* obtained with the various methods are compared. No significant differences in any of the time course metrics were detected between vPIVOT and PIVOT or OxFlow. Figure 3 shows the average vPIVOT-derived parameters during the dynamic exercise paradigm. Figure 4 shows the average perfusion response to exercise in each of the muscles in the leg. $$$\dot{V}O_{2}$$$ was quantified based on perfusion or arterial flow and average results are displayed in Figure 5. The total integrated $$$\dot{V}O_{2}$$$ response is very similar when comparing $$$\dot{V}O_{2}$$$ calculated with perfusion or arterial flow, however the response dynamics differ substantially.

Slice thickness and flip angle were reduced in the superior GRE acquisition to minimize its potential impact on the quantification of perfusion. Results in Figure 3 show that vPIVOT is sensitive to the changes in oxygen saturation and microvascular and macrovascular blood flow during the transition from exercise to rest, known as the off-kinetics of exercise. Plantar flexion contraction recruits the posterior compartment of the leg, which comprises the soleus and gastrocnemius muscles. Perfusion data in Figure 4 shows hyperperfusion in the gastrocnemius, with a slight increase in the soleus, however no activation was observed in the anterior compartment. The relationship between micro and macrovascular blood flow differs substantially between post-ischemia reactive hyperemia and post-exercise active hyperemia. This may be due to the difference in oxygen demand of the tissue, which does not increase in reactive hyperemia¹¹.

vPIVOT can be used to quantify $$$\dot{V}O_{2}$$$ dynamics and may be useful in separating the microvascular and macrovascular contributions to the hyperemic response.