Estimation of intersubject variability of cerebral blood flow measurements using MRI and positron emission tomography

Abstract

To investigate the within and between subject variability of quantitative cerebral blood flow (CBF) measurements in normal subjects using various MRI techniques and positron emission tomography (PET). Repeated CBF measurements were performed in 17 healthy, young subjects using three different MRI techniques: arterial spin labeling (ASL), dynamic contrast enhanced T1 weighted perfusion MRI (DCE) and phase contrast mapping (PCM). All MRI measurements were performed within the same session. In 10 of the subjects repeated CBF measurements by (15) O labeled water PET had recently been performed. A mixed linear model was used to estimate between subject (CV(betw)) and within subject (CV(with)) coefficients of variation. Mean global CBF, CV(betw) and CV(with) using each of the four methods were for PCM 65.2 mL/100 g/min, 17.4% and 7.4%, for ASL 37.1 mL/100 g/min, 16.2% and 4.8%, for DCE 43.0 mL/100 g/min, 20.0%, 15.1% and for PET 41.9 mL/100 g/min, 16.5% and 11.9%, respectively. Only for DCE and PCM a significant positive correlation between measurements was demonstrated. These findings confirm large between subject variability in CBF measurements, but suggest also that in healthy subjects a subject-method interaction is a possible source of between subject variability and of method differences.