Cardiac Application Tips

Cardiac Viability Imaging Using Inversion Time Scout

Contributor: Vivian Lee, M.D., Ph.D. and Manmeen Kaur, M.D.

Introduction

Delayed contrast-enhanced magnetic resonance imaging (MRI) is becoming an increasingly accepted modality for the diagnosis of myocardial infarction. The imaging technique is premised on the delayed enhancement and washout of extracellular contrast agents, such as gadolinium chelates, by infarcted myocardium relative to viable myocardium. About 10 – 30 min after injection of contrast material, infarcted myocardium will appear slightly hyper-enhanced relative to uninfarcted myocardium, however this difference may be difficult to detect on routine T1-weighted imaging (Figure 1). A critical advance in this field has been the implementation of an inversion-recovery T1-weighted approach whereby the time between the 180° inversion pulse and RF excitation, referred to as the inversion time (TI), is selected to null signal from uninfarcted myocardium. With uninfarcted or viable myocardium nulled, the hyper-enhancing infarct becomes conspicuous.

The selection of an optimal TI (TI0) to null the uninfarcted myocardium can critically affect the diagnostic value of the infarct images because the TI determines the relative signal intensity of the infarcted and uninfarcted myocardium. Choosing the TI can be difficult because differences in the contrast dose, patient body habitus, cardiac function, and time after contrast administration can cause the optimal TI to vary considerably. Typically a TI in the range of 200-350 msec is appropriate for images obtained 10 – 20 minutes after administration of about 0.1 to 0.2 mmol per kg of body weight of gadolinium chelate. Nevertheless, because the optimal TI is not known a priori, a series of trial-and-error guesses at TI intervals of 25 – 50 msec usually precedes successful infarct imaging.

This application tip describes a new cine inversion recovery segmented k-space true fast imaging with steady-state precession (FISP) sequence that helps to determine the optimal TI for infarct imaging in a single breath-hold acquisition.

Inversion Time (TI) Scout Sequence

The sequence is an inversion recovery segmented-k-space cine gradient echo sequence. Following an electrocardiographic trigger, an inversion pulse (180°) is applied. Then a true FISP readout is implemented with a segmented k-space approach, where each segment corresponds to a different inversion time, so that the acquired cine images trace the longitudinal relaxation of the tissues imaged (Figure 2). Because images are gated to the electrocardiogram, consecutive images also depict systolic and diastolic changes in cardiac morphology during the cardiac cycle. Segments are centrically reordered to reduce sensitivity of image contrast to heart beat variation. Also, preparation pulses are performed before the initial acquisition to reduce oscillatory magnetization for steady state imaging. Data is collected during every other heartbeat to allow for adequate longitudinal relaxation.

In one implementation (Figures 3 and 4) (1.5T Magnetom Symphony with Quantum gradients), sequence parameters are as follows: TR (msec)/TE (msec)/flip angle = 2.9/1.3/50º, matrix 104 x 192, 15 lines/segment, field of view (FOV) 263 x 350 mm, slice thickness 6 mm, temporal resolution 43 msec, acquisition time = 14 heartbeats. A total of 16-20 images are obtained, whereby of the first image = 100 msec and then TI increases at 43 msec intervals. The optimal TI, TI0, was defined visually as the inversion time at which the uninfarcted myocardium was nulled (Figures 1 and 2). Note that the TI0 could be determined regardless of the presence or absence of infarct on the TI mapping images.

Infarct imaging is then performed using the selected TI0 and a 2D segmented k-space inversion recovery true FISP sequence with typical parameters: TR/TE/FA = 700/1.4/50°, 33 lines/segment, matrix 129 x 256, FOV 262 x 350, acquisition time for 3 slices = 12 heartbeats. Images (8 mm with 2 mm gap) were acquired in the short axis from the left ventricular base to the apex as well as in the horizontal and vertical long axis planes (Figure 5).

Figure 1: Differential enhancement and washout patterns of normal (viable) versus infarcted myocardium after gadolinium (GD) contrast injection. About 10 – 30 min after injection of contrast material, infarcted myocardium will appear slightly hyper-enhanced relative to uninfarcted myocardium.

Figure 2: Pulse sequence diagram for inversion recovery true FISP sequence for inversion time (TI) mapping. Following the ECG trigger, an inversion recovery pulse is applied. Then a true FISP readout is performed with a segmented k-space approach, where each centrically-encoded segment (15 lines in this implementation) corresponds to a different inversion time, so that the acquired cine images (inversion time interval, DTI = 43 msec) trace the longitudinal relaxation of the tissues. Preparation pulses before the initial acquisition reduce oscillatory magnetization for steady state imaging. Data are collected during every other heartbeat to allow for adequate longitudinal relaxation. The optimal time to inversion (TI0) was defined visually as the inversion time at which the uninfarcted myocardium was nulled.

Figure 3: Plot of signal intensity (SI) versus inversion time (TI) for myocardial infarct (green), uninfarcted myocardium (blue) for a delayed contrast-enhanced inversion recovery T1-weighted sequence. Because the magnitude data are collected, the recovery curve, in fact, has the appearance of the data points above the x-axis, creating ambiguity in the image contrast depending on the TI used. Example images corresponding to each inversion time are shown below with areas of infarct (arrows) optimally visualized at TI0, where uninfarcted myocardium is nulled. In this example, at TI-3, the image contrast is reversed, with infarcted myocardium appearing nulled (arrows), and uninfarcted myocardium bright. At TIs longer than TI0, the boundary between infarct and uninfarcted myocardium becomes less well defined.

TI-3 = 170 ms TI-2 = 203 ms TI-1 = 238 ms   TI0 = 275 ms   TI+1 = 308 ms TI+2 = 342 ms TI+3 = 375 ms

(in a separate window) Figure 5b Figure 5c

Figure 5. Inversion recovery true FISP infarct images in same subject whose TI mapping was illustrated in Figure 3. (A) Short axis cine of 2D segmented k-space inversion recovery true FISP sequence using selected TI0 from the left ventricular base to the apex. (B) Horizontal and (C) vertical long axis images performed 10 – 15 min after injection of 0.15 mmol/kg gadolinium contrast material using TI = 229 msec, corresponding to TI0 determined from the TI mapping sequence. Images show subendocardial infarct (arrows on 5B and 5C) appearing slightly hyperintense relative to blood pool, and markedly hyperintense relative to the nulled uninfarcted myocardium.

References:

1. Gupta, A. et al. 2003. Delayed Contrast-Enhanced Magnetic Resonance Imaging for Myocardial Infarction: Optimizing Inversion Times. Abstract presented to RSNA 2003 89th Scientific Assembly and Annual Meeting. Chicago, Illinois, Nov 30 – Dec 5 2003.