Files in this Data Supplement:

• Addendum - (PDF) (268 KB)
• Figure I - (TIFF) (3.04 MB) Quantitative myocardial perfusion at rest and during adenosine vasodilation in the same patient. Top figure demonstrates the signal intensity (SI) curves (arbitrary units=a.u.) during the first pass of a contrast agent bolus in the hypertrophied basal myocardial segment (marked as *), and the bottom figure represents the SI curves in the blood pool used as the arterial input function.
• Figure II - (TIFF) (131 KB) The circles show the relation between signal intensity, measured with a saturation-recovery prepared turbo FLASH sequence, and the relaxation rate in saline-filled phantoms doped with Gd-DTPA contrast agent. The R1 relaxation rate constant of the phantoms was determined with a spin-echo sequence. The evolution of the longitudinal magnetization of flowing blood, imaged with a non-slice-selective saturation recovery prepared turbo FLASH acquisition, was simulated using the approach of Peeters et al2 and assuming a linear phase encoding order. The blue solid line shows the simulation result for stationaery fluid, while for the green curve a velocity of v=10cm/s was assumed with otherwise identical parameters. The slice thickness was equal to 8 mm in the simulations. The dotted line represents a linear extrapolation from the region of low R1 values, and corresponds to the assumption of strict linear proportionality between signal intensity and signal, which was used for the analysis of the perfusion studies.
• Figure III - (TIFF) (166 KB) Left: The value of the smoothness side constraint for the impulse response is evaluated for a range of the regularization parameter lambda and plotted against the sum of the residuals (magnitude) evaluated for the same values of lambda, with the corresponding impulse response. The shape has an L-shaped knee, and the corresponding value of lambda represents an optimal choice for the regularization parameter. Right: The impulse response, shown here for rest (top) and hyperemia (bottom), was represented as a sum of B-splines. This B-spline representation was chosen to improve the numerical stability of the deconvultion process. The tissue impulse response curves shown here correspond to the optimal choices of the regularization parameter. The blood flow is estimated from the amplitude of the impulse response.