Infarct density distribution by MRI in the porcine model of acute and chronic myocardial infarction as a potential method transferable to the clinic

Academic Article


  • To study the feasibility of a myocardial infarct (MI) quantification method [signal intensity-based percent infarct mapping (SI-PIM)] that is able to evaluate not only the size, but also the density distribution of the MI. In 14 male swine, MI was generated by 90 min of closed-chest balloon occlusion followed by reperfusion. Seven (n = 7) or 56 (n = 7) days after reperfusion, Gd-DTPA-bolus and con-tinuous- infusion enhanced late gadolinium enhancement (LGE) MRI, and R1-mapping were carried out and post mortem triphenyl- tetrazolium-chloride (TTC) staining was performed. MI was quantified using binary [2 or 5 standard deviation (SD)], SI-PIM and R1-PIM methods. Infarct fraction (IF), and infarct-involved voxel fraction (IIVF) were determined by each MRI method. Bias of each method was compared to the TTC technique. The accuracy of MI quantification did not depend on the method of contrast administration or the age of the MI. IFs obtained by either of the two PIM methods were statistically not different from the IFs derived from the TTC measurements at either MI age. IFs obtained from the binary 2SD method overestimated IF obtained from TTC. IIVF among the three different PIM methods did not vary, but with the binary methods the IIVF gradually decreased with increasing the threshold limit. The advantage of SI-PIM over the conventional binary method is the ability to represent not only IF but also the density distribution of the MI. Since the SI-PIM methods are based on a single LGE acquisition, the bolus-data-based SI-PIM method can effortlessly be incorporated into the clinical image post-processing procedure. © Springer Science+Business Media 2014.
  • Digital Object Identifier (doi)

    Author List

  • Varga-Szemes A; Simor T; Lenkey Z; Van Der Geest RJ; Kirschner R; Toth L; Brott BC; Elgavish A; Elgavish GA
  • Start Page

  • 937
  • End Page

  • 948
  • Volume

  • 30
  • Issue

  • 5