Skip to content


  • Poster presentation
  • Open Access

Inline non-rigid motion-corrected t2 mapping of myocardium

  • 1,
  • 2,
  • 3,
  • 4,
  • 1,
  • 2,
  • 3,
  • 1 and
  • 1
Journal of Cardiovascular Magnetic Resonance201012 (Suppl 1) :P229

  • Published:


  • Motion Correction
  • Free Breathing
  • Typical Imaging Parameter
  • SSFP Readout
  • Typical Processing Time


Myocardial T2 is altered in certain pathologies like acute coronary syndrome (ACS), myocarditis and allograft rejection, which have traditionally been detected using T2-weighted (T2W) techniques. These T2W techniques suffer from several drawbacks [1], most of which can be addressed by a quantitative T2 mapping approach. Single-shot T2-prepared SSFP (T2prep-SSFP) has been proposed for T2 Mapping [2]; however, image mis-registration due to failed breath-hold or inconsistent cardiac rhythm degrades the accuracy of pixel-wise curve fitting. We propose a comprehensive T2 mapping technique that includes automatic registration of T2prep images with an inline implementation to enable clinical application.


To design and implement a rapid motion corrected inline T2 mapping technique for myocardium.


Three T2prep times were used (Table 1). The non-selective T2prep pulse combined with a fast SSFP readout makes each T2P acquisition insensitive to motion. To correct for motion between images, a fast variational non-rigid registration algorithm [3] was employed which aligned all T2prep frames to the center frame. T2 maps were then computed by fitting intensities of corrected images to the mono-exponential decay curve: Signal = M0 * exp(-T2prep/T2). Finally, color-coded T2maps were generated inline; the entire post-processing is unsupervised and typical processing time was less than 1 s.

T2 maps were acquired from a mid-ventricular short-axis (SAX) slice in each of 5 healthy subjects using breath-hold and free breathing to assess the performance of motion correction. Typical imaging parameters are listed in Table 1. Additionally, to demonstrate the ability of the technique to detect edema, T2Map, STIR and LGE images were acquired in a canine model of acute myocardial infarction (AMI) and in a patient diagnosed with ACS.

Table 1

Imaging Parameters

Parameter name


Number of T2Preps


T2P times

0 (no prep), 24, 55 ms

Acquisition window

~ 150 ms

Total acquisition time

9 RR intervals

Acquisition mode

Single shot


2.5 × 2.75 × 8 mm

Parallel acquisition/acceleration/reference lines


Flip angle

70 degrees

Field strength

1.5 T


T2 of normal human myocardium was measured to be 56.1 ms. T2 values did not show any significant difference with and without free breathing (p-value = 0.55). Figure 1 shows a T2 map acquired with (a) and without (b) breath-hold and after applying motion correction to free-breathing images (c). Ischemic region had a significantly higher T2 than the remote region in dog (65 vs. 46.8 ms) and in patient (69.6 vs. 48.4 ms), shown in figures 2 and 3 respectively.
Figure 1
Figure 1

T2 Maps acquired with breath-hold (A), without breath-hold (B&C). The motion-correctionis applied only in C. Note the motion-related (loss of pixels) artifact in B (arrow) whereas in C, the motion correction has corrected for this.

Figure 2
Figure 2

Canine model of AMI. T2 Map (A), Short-inversion Time Inversion Recovery (STIR) (B) and Late Gadolinium Enhancement (LGE) (C) images showing the ischemia and infarct (arrows) in the anterior region (perfusion bed of LAD).

Figure 3
Figure 3

T2 Map (A), STIR (B) and LGE (C) images showing the ischemia and infarct (arrows in the patient dianosed with Acute Coronary Syndrom.


We have demonstrated a rapid T2 Mapping technique with integrated unsupervised motion correction and inline colormap generation.

Authors’ Affiliations

The Ohio State University, Columbus, OH, USA
Siemens Corporate Research, Princeton, NJ, USA
Siemens Healthcare, Chicago, IL, USA
Northwestern University, Chicago, IL, USA


  1. Circulation. 2008, 118 (8): 795-6. 10.1161/CIRCULATIONAHA.108.797373.Google Scholar
  2. Magn Reson Med. 2007, 57 (5): 960-6. 10.1002/mrm.21208.Google Scholar
  3. Med Image Comput Comput Assist Interv Int Conf Med Image Comput Comput Assist Interv. 2008, 11 (Pt 2): 35-43.Google Scholar


© Giri et al; licensee BioMed Central Ltd. 2010

This article is published under license to BioMed Central Ltd.