INTRODUCTION: The investigation of pulmonary perfusion by three-dimensional (3D) dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) was proposed recently. Subtraction images are generated for clinical evaluation, but temporal information is lost and perfusion defects might therefore be masked in this process. The aim of this study is to demonstrate a simple analysis strategy and classification for 3D-DCE-MRI perfusion datasets in the lung without omitting the temporal information. MATERIALS AND METHODS: Pulmonary perfusion measurements were performed in patients with different lung diseases using a 1.5 T MR-scanner with a time-resolved 3D-GRE pulse sequence. 25 3D-volumes were acquired after iv-injection of 0.1 mmol/kg KG Gadolinium-DTPA. Three parameters were determined for each pixel: (1) peak enhancement S(n,max) normalized to the arterial input function to detect regions of reduced perfusion; (2) time between arterial peak enhancement in the large pulmonary artery and tissue peak enhancement τ to visualize regions with delayed bolus onset; and (3) ratio R=S(n,max)/τ was calculated to visualize impaired perfusion, irrespectively of whether related to reduced or delayed perfusion. RESULTS: A manual selection of peak perfusion images is not required. Five different types of perfusion can be found: (1) normal perfusion; (2) delayed non-reduced perfusion; (3) reduced non-delayed perfusion; (4) reduced and delayed perfusion; and (5) no perfusion. Types II and IV could not be seen in subtraction images since the temporal information is necessary for this purpose. CONCLUSIONS: The analysis strategy in this study allows for a simple and observer-independent visualization and classification of impaired perfusion in dynamic contrast-enhanced pulmonary perfusion MRI by using the temporal information of the datasets.
INTRODUCTION: The investigation of pulmonary perfusion by three-dimensional (3D) dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) was proposed recently. Subtraction images are generated for clinical evaluation, but temporal information is lost and perfusion defects might therefore be masked in this process. The aim of this study is to demonstrate a simple analysis strategy and classification for 3D-DCE-MRI perfusion datasets in the lung without omitting the temporal information. MATERIALS AND METHODS: Pulmonary perfusion measurements were performed in patients with different lung diseases using a 1.5 T MR-scanner with a time-resolved 3D-GRE pulse sequence. 25 3D-volumes were acquired after iv-injection of 0.1 mmol/kg KG Gadolinium-DTPA. Three parameters were determined for each pixel: (1) peak enhancement S(n,max) normalized to the arterial input function to detect regions of reduced perfusion; (2) time between arterial peak enhancement in the large pulmonary artery and tissue peak enhancement τ to visualize regions with delayed bolus onset; and (3) ratio R=S(n,max)/τ was calculated to visualize impaired perfusion, irrespectively of whether related to reduced or delayed perfusion. RESULTS: A manual selection of peak perfusion images is not required. Five different types of perfusion can be found: (1) normal perfusion; (2) delayed non-reduced perfusion; (3) reduced non-delayed perfusion; (4) reduced and delayed perfusion; and (5) no perfusion. Types II and IV could not be seen in subtraction images since the temporal information is necessary for this purpose. CONCLUSIONS: The analysis strategy in this study allows for a simple and observer-independent visualization and classification of impaired perfusion in dynamic contrast-enhanced pulmonary perfusion MRI by using the temporal information of the datasets.
Authors: Patricia Leutz-Schmidt; Monika Eichinger; Mirjam Stahl; Olaf Sommerburg; Jürgen Biederer; Hans-Ulrich Kauczor; Michael U Puderbach; Marcus A Mall; Mark O Wielpütz Journal: Radiologe Date: 2019-12 Impact factor: 0.635
Authors: Jürgen Biederer; S Mirsadraee; M Beer; F Molinari; C Hintze; G Bauman; M Both; E J R Van Beek; J Wild; M Puderbach Journal: Insights Imaging Date: 2012-01-15
Authors: Jimmy Z Yu; Tobias Granberg; Roya Shams; Sven Petersson; Magnus Sköld; Sven Nyrén; Johan Lundberg Journal: J Intern Med Date: 2022-08-10 Impact factor: 13.068