Skip to main content
Figure 6 | Journal of Cardiovascular Magnetic Resonance

Figure 6

From: Quantification of myocardial perfusion by cardiovascular magnetic resonance

Figure 6

Deconvolution of the measured myocardial signal intensity curve (blue circles) with the arterial input (b), was performed with a Fermi function model of the impulse response (c). Deconvolution of the measured myocardial signal intensity curve (blue circles) with the arterial input shown in (b), was performed with a Fermi function model of the impulse response as in (c). The time window for the fit excluded the portion of the signal intensity curve after the first pass in the blood pool (grayed-out). Also the initial transition of the myocardial signal to a steady state level, before injection of contrast was excluded (grayed out). The best fit to the measured data, shown as solid blue line in (a) was obtained by non-linear least-squares fitting, using a Marquardt-Levenberg algorithm in the Matlab environment (Mathworks, Natick, MA). The fitting function is calculated by convolution of the Fermi-function with the measured arterial input. The impulse response shape corresponding to this best fit is shown in panel (c). The myocardial blood flow is estimated from the initial amplitude of the Fermi impulse response.

Back to article page