Skip to content


  • Oral presentation
  • Open Access

Detecting acute reperfusion myocardial hemorrhage with CMR: a translational study

  • 1, 2,
  • 3,
  • 2 and
  • 2
Journal of Cardiovascular Magnetic Resonance201214 (Suppl 1) :O61

  • Published:


  • Image Resolution
  • Translational Study
  • Late Enhancement
  • Noninvasive Detection
  • Monoexponential Decay


Hemorrhage is a frequent hallmark of large acute reperfused myocardial infarctions (rMI). Recently, there has been a growing interest in CMR for noninvasive detection of hemorrhage in rMI. While T2*- and T2-weighted CMR have been used for this purpose, there is a lack of consensus on which of these methods is the most appropriate and reliable. We investigated the relative abilities of T2 and T2* CMR for detecting hemorrhage in rMI in a canine model and patients.


Canines (n=14), subjected to ischemia-reperfusion (I/R) injury (3 hrs of LAD occlusion followed by reperfusion), underwent CMR (1.5T) studies on day 5 post reperfusion. T2*-weighted (multi GRE; TE=3.4-18.4ms (6 echoes)), T2-weighted (T2-prep SSFP; prep times=0, 24 and 55 ms), T2-STIR (TE=64 ms) and Late Enhancement (LE) images covering the LV were acquired. Imaging resolution of all the scans was 1.3x1.3x8 mm3.

Patients (n=14) underwent CMR (1.5T) on day 3 post angioplasty for STEMI after providing informed consent. T2*-weighted (TE=2.6-13.8ms (6 echoes)), T2-STIR (TE=61ms) and LE images covering the LV were acquired. Imaging resolution of all the scans was 1.4x1.4x10 mm3.

T2* and T2 maps were constructed by fitting the multi-echo data to monoexponential decay. A threshold-based signal analysis was used to identify hemorrhagic (Hemo+) and non-hemorrhagic (Hemo-) infarcts. T2-STIR signal intensity (STIR-SI), T2* and T2 values, measured from Remote, Hemo- and Hemo+ regions, were compared. Statistical significance was set at p<0.05.


Representative T2* and T2 maps, and T2-STIR and LE images (acquired from a canine on day 5 post I/R injury) are shown in Fig. 1A. Mean T2* of Hemo+ was lower than both Remote and Hemo- regions (-39%; p<0.001; Fig. 1B and 1C), while no differences were observed in T2* between Remote and Hemo- (p=0.27). Compared to Remote, mean T2 of both Hemo- and Hemo+ regions were elevated (26% and 17% respectively; p<0.001), with T2 of Hemo- greater than T2 of Hemo+ (p<0.001). A similar trend was observed in T2-STIR images as well; STIR-SI of Hemo+ and Hemo- were greater than Remote (59% and 31% respectively; p<0.001), while STIR-SI of Hemo- was greater than Hemo+ (p<0.001).
Figure 1
Figure 1

(A) Representative set of T2* (color-coded) and T2 maps, T2-STIR and LE images acquired from a canine on day 5 post reperfusion are shown. Arrows point to the site of infarction on LE image. Hemorrhagic territory (Hemo+) is enclosed in blue ROI, and remote territory is enclosed in red ROI. (B) Mean T2* of Hemo+ was significantly lower than those of Remote and Hemo- (#, p<0.001). Mean T2 of both Hemo+ and Hemo- were elevated compared to Remote (* and ^, p<0.001), with T2 of Hemo- higher than that of Hemo+ (p<0.001). (C) Compared to remote, T2* of Hemo+ decreased by 39% (p<0.001), while T2* of Hemo- remained unchanged (p=0.21). T2 of Hemo- and Hemo+ remained elevated by 26% and 17% respectively (p<0.001 for both cases). STIR-SI of Hemo- and Hemo+ were also elevated by 59% and 31% respectively (p<0.001 for both cases).

Representative T2* map, T2-STIR and LE images, acquired from a patient (day 3 post angioplasty) are shown in Figure 2. Consistent with the animal studies, mean T2* of Hemo+ was lower than the mean T2* of both Remote and Hemo- (-46%, p<0.001; Fig.2), while no differences were observed between T2* of Remote and Hemo- (p=0.61). Mean STIR-SI of both Hemo- and Hemo+ were greater than that of Remote (78% and 33% respectively; p<0.001), with the mean STIR-SI of Hemo- greater than that of Hemo+ (p<0.001).


This work was supported in part by grants from American Heart Association (SDG 0735099N) and National Heart, Lung, And Blood Institute (HL091989).

Authors’ Affiliations

Biomedical Engineering, University of California, Los Angeles, Los Angeles, CA, USA
Biomedical Sciences, Cedars-Sinai Medical Center, Los Angeles, CA, USA
Québec Heart and Lung Institute, Laval University, Québec City, QC, Canada, USA


© Kali et al; licensee BioMed Central Ltd. 2012

This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.