Meeting Banner
Abstract #1220

Voxel-Wise Quantitative Assessment of Myocardial Perfusion: A Comparison of Four Different Deconvolution Algorithms using Real Flow Values

Niloufar Zarinabad Nooralipour1, Amedeo Chiribiri1, Gilion Hautvast2, Aruna Vishnu Arujuna1, Eike Nagel1, Philip Batchelor3

1The Centre of Excellence in Medical Engineering, Kings College London, London, United Kingdom; 2Imaging Systems- MR, Philips Healthcare, Netherlands; 3The Centre of Excellence in Medical Engineering, Kings College London, London, United Kingdom


In this study four different deconvolution algorithms have been applied to voxel-wise analysis of first-pass myocardial perfusion MR data. We aimed to test robustness of these methods to noise and to evaluate the accuracy of the perfusion estimates by validating the estimated perfusion values in a hardware perfusion phantom with true perfusion values, measured by means of precision flow-meters. This study demonstrated that Auto-Regressive Moving Average model is the least sensitive method to noise, while achieving almost similar accuracy to exponential basis deconvolution, which is the superior model in terms of accuracy of estimation, and compares favourably over the Fermi function modelling quantification method