Kerstin Ulrike Bauner MD, Presenter: Nothing to Disclose

Andreas Biffar, Abstract Co-Author: Nothing to Disclose

Daniel Theisen MD, Abstract Co-Author: Nothing to Disclose

Andreas Greiser PhD, Abstract Co-Author: Employee, Siemens AG

Maximilian F. Reiser MD, Abstract Co-Author: Nothing to Disclose

Bernd J. Wintersperger MD, Abstract Co-Author: Speakers Bureau, Bayer AG Speakers Bureau, Siemens AG

To evaluate the diagnostic accuracy of unenhanced and enhanced myocardial T1 mapping cardiac MR in the detection of chronic myocardial infarction (CMI).

26 patients with CMI were examined at 1.5T (Avanto, Siemens Healthcare). A modified Look-Locker inversion recovery (MOLLI) sequence (TR/TE 200.7/1.03msec; TI 100-4000msec; flip 35°) was performed prior and 10min post-contrast (0.15mmol/kg Gadobutrol, Bayer Healthcare) at an apical, midmyocardial and basal short axis position. 15 minutes post-contrast IR GRE imaging was performed for assessment of Late Gadolinium Enhancement (LGE) . Data were post-processed on an external workstation with in-house built software (PMI 0.4) and T1-maps of unenhanced/ enhanced MOLLI-data calculated. CMI on T1 maps were defined as areas with T1-values >3SD different as compared to normal remote myocardium. On LGE imaging (IR-GRE) CMI size was defined as areas with SI >2SD different than normal remote myocardium on matching slice locations. The presence of CMI was assessed according to the 16 segment model and compared using ROC analyses. The infarction sizes were compared with a Bland-Altman analysis.

Unenhanced T1 maps detected CMI with a sensitivity of 75.8% and a specificity 99.7% with a positive and negative predictive value (PPV and NPV) of 99.0% and 91.1% while contrast enhanced T1 maps demonstrated a sensitivity and specificity of of 95.8% and 99.3% respectively with a PPV and NPV of 98.2% and 98.3% . Bland-Altman analysis revealed underestimation of CMI size in unenhanced T1 maps as compared to LGE with a systematic error of 1cm² (limit of agreement, LoA -3.9 – 1.8) while contrast enhanced T1maps only demonstrated a systematic error of 0.3cm² (LoA -1.6 – 1.0).

CMI can be detected with high diagnostic accuracy on contrast enhanced T1-maps, while in unenhanced examinations CMI may be missed and CMI sizes may be underestimated.

Imaging CMI on the basis of contrast-enhanced T1-maps may be used instead of IR sequences, which offers the advantage of being independent of the selection of the correct inversion time.

Bauner, K, Biffar, A, Theisen, D, Greiser, A, Reiser, M, Wintersperger, B, Myocardial T1 Mapping vs Late Gadolinium Enhancement in Assessment of Chronic Myocardial Infarction.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11034335.html