Mona Salehi Ravesh1, Michael Puderbach2,
Sebastian Ley3, Julia Ley-Zaporzhan3, Frank Risse1,
Wilfried Schranz4, Wolfhard Semmler1, Frederik Bernd
Laun1
1Department of Medical Physics in
Radiology, German Cancer Research Center, Heidelberg, Germany; 2Department
of Radiology, German Cancer Research Center, Heidelberg, Germany; 3Diagnostic
and Interventional Radiology, University Hospital Heidelberg, Heidelberg,
Germany; 4Nonlinear Physics Group, Faculty of Physics, University
of Vienna, Vienna, Austria
Lung
perfusion is a crucial prerequisite for effective gas exchange. An accurate
quantification of pulmonary perfusion is therefore important for diagnostic
considerations and treatment planning in various diseases of the lungs.The
assessment of pulmonary perfusion by Dynamic Contrast-Enhanced Magnetic
Resonance Imaging requires deconvolution of the arterial input function. In
the presence of noise this is an ill-posed problem which leads to strongly
oscillating, unphysical solutions when it is solved without regularization.
In this study a novel method to quantify the pulmonary perfusion is used and
compared to the singular value decomposition and L-curve criterion based on simulated
and patient data.