Abstract
A novel pulse sequence scheme is presented that allows the measurement and mapping of myocardial T 1 in vivo on a 1.5 Tesla MR system within a single breath-hold. Two major modifications of conventional Look-Locker (LL) imaging are introduced: 1) selective data acquisition, and 2) merging of data from multiple LL experiments into one data set. Each modified LL inversion recovery (MOLLI) study consisted of three successive LL inversion recovery (IR) experiments with different inversion times. We acquired images in late diastole using a single-shot steady-state free-precession (SSFP) technique, combined with sensitivity encoding to achieve a data acquisition window of <200 ms duration. We calculated T 1 using signal intensities from regions of interest and pixel by pixel. T 1 accuracy at different heart rates derived from simulated ECG signals was tested in phantoms. T 1 estimates showed small systematic error for T 1 values from 191 to 1196 ms. In vivo T 1 mapping was performed in two healthy volunteers and in one patient with acute myocardial infarction before and after administration of Gd-DTPA. T 1 values for myocardium and noncardiac structures were in good agreement with values available from the literature. The region of infarction was clearly visualized. MOLLI provides high-resolution T 1 maps of human myocardium in native and post-contrast situations within a single breath-hold. Despite recent technological advances, in vivo T 1 quantification of the myocardium with modern magnetic resonance (MR) systems remains a challenge because of severe time constraints due to cardiac and respiratory motion. While myocardial T 1 is shorter and therefore relatively easier to measure at low field strengths, it has a value of ϳ1000 ms at a field strength of 1.5 T, exceeding the duration of the cardiac cycle (ϳ600 -1200 ms) in most subjects (1, 2) . Since standard inversion recovery (IR) measurements require a relaxation period of four to five times T 1 to allow for full magnetization recovery after each 180°pulse (3) , only four to five such single-point IR experiments can be performed within one breath-hold (ca. 20 s). To achieve accurate T 1 estimates from a three-parameter curve-fitting procedure, as is commonly employed, data from at least six to 10 time points should be available (4) . The multipoint approach, as first described by Look and Locker (5) , samples the relaxation curve multiple times after an initial preparation pulse (6) . This technique has been shown theoretically to be highly efficient (7) , and has been widely used for T 1 measurements of the brain (8 -11) . It is not suitable for pixel-by-pixel T 1 mapping of the heart because data acquisition is performed continuously throughout the cardiac cycle without regard for cardiac motion, which means that T 1 values can only be derived for regions of interest (ROIs) that must be defined manually for every frame (1) . The resultant T 1 values may consequently be subject to inaccuracy caused by misregistration effects. In this work we present a pulse sequence scheme that allows for accurate in vivo T 1 measurements and T 1 mapping of myocardium with high spatial resolution and within a single breath-hold. To overcome the limitations of the conventional LL approach for cardiac applications, we propose a modified LL IR scheme (MOLLI), which introduces two principles to the standard LL sequence: 1) selective data acquisition at a given time of the cardiac cycle over successive heartbeats, and (2) merging of image sets from multiple LL experiments with varying inversion times (TIs) into one data set. While selective data acquisition effectively decreases the number of images acquired in each LL experiment to one per heartbeat, the use of multiple LL experiments with different TIs increases the number of samples of the relaxation curve to a value that is sufficiently high for accurate T 1 estimation. The IR preparation pulse is used to yield the maximum dynamic range of the signal. A balanced steady-state free precession (SSFP) readout is chosen over conventional gradient-echo (GE) readout because of its higher signal-to-noise ratio (SNR) and lower tendency to modulate the relaxation curve (12) . To minimize artifacts from cardiac motion, the image data acquisition window is restricted to Ͻ200 ms in end-diastole by the use of sensitivity encoding (SENSE) (13) with a reduction factor of 2. We investigated the accuracy of T 1 measurements with MOLLI using gel phantoms for a wide range of heart rates derived from simulated electrocardiogram (ECG) signals. T 1 maps and the T 1 values derived from in vivo experiments in two healthy volunteers and one patient with acute myocardial infarction are presented.
Showing the abstract — retrieve the full paper via the Exa API.