Adaptive Sequential Design for Optimal Scheduling of Continuous ASL Data Acquisition

Introduction: It has been demonstrated that Optimal Sampling Schedule (OSS) theory [1] can be used as an objective strategy for choosing a TI sampling scheme in pulsed arterial spin labelling (PASL) experiments, yielding an improvement in parameter estimation precision [2]. However, this requires a priori estimates of the likely value of the fitted parameters. Adaptive Sequential Design (ASD) addresses this issue by incorporating the OSS strategy into an adaptive process which iteratively updates the parameter estimates and adjusts the sampling schedule accordingly as data are acquired. Simulation work on an ASD-CASL protocol demonstrated that an optimized sampling strategy approach can be tuned in real time to incorporate pathological/abnormal parameter information, thereby improving parameter estimation precision [3]. Here we implemented the ASD concept with real-time feedback on a Siemens MR scanner. Good parameter estimations were achieved on-line in normal volunteers.