Use of left ventricle blood pool oxygenation-sensitive signal intensity as a measure of arterial hemoglobin saturation

Background Oxygenation-sensitive cardiovascular magnetic resonance (OS-CMR) relies on the attenuating effects of deoxyhemoglobin as an intrinsic contrast. When using OS-CMR to assess myocardial oxygenation, failing to account for effects on myocardial oxygen-sensitive signal intensity (OS-SI) due to arterial hemoglobin desaturation could confound results attributed to changes in perfusion and tissue metabolism. In particular, this can be relevant when using breath-holds as a vasodilating mechanism. Using an animal model we first assessed if arterial saturation can be non-invasively measured using the blood pool signal (SI) in OS-CMR images, and secondly if this effect is relevant for human studies. Methods In eight anaesthetized healthy swine, cannulas were placed in the femoral artery for blood sampling. Arterial and venous blood samples were taken from a human cohort of 36 healthy humans and 29 obstructive sleep apnea syndrome (OSAS) patients. All subjects performed a breathing maneuver of 60s hyperventilation followed by a long end-expiration breath-hold. In humans, the breath-hold duration was voluntary and analyzed for 30s, and in animals, manual ventilation performed and paused for 90s. OS-CMR images were acquired continuously throughout the breath-hold in a mid-slice short-axis slice with a bSSFP sequence (FA=35°, TR=3.4 ms) at 3T. SI values were captured in the left ventricle blood pool at end systole and expressed as a change over the breath-hold.