Limitations of a multiplicative correction of partial volume effects on the arterial input function in bolus-tracking perfusion

Introduction: Quantification of cerebral blood flow (CBF) using bolus-tracking MRI is critically dependent on correct determination of the arterial input function (AIF) (1,2). The AIF is the concentration of tracer which enters the brain as a function of time. It is sampled in a supplying artery, which is often subject to a partial volume effect (PVE) due to the limited spatial resolution resulting from the requirement of a short image sampling time. A commonly used correction approach entails a multiplicative scaling of a measured AIF concentration-time curve (e.g. (3-5)). Here we show using in vivo data obtained with a T1-weighted perfusion imaging approach (6) that such a multiplicative rescaling can lead to distortion of the AIF resulting in serious errors in CBF estimates. We propose to use an easily measurable quantity denoted the tissue signal fraction (TSF) as a measure of the applicability of a multiplicative rescaling.