ISMRM 23rd Annual Meeting & Exhibition • 30 May - 05 June 2015 • Toronto, Ontario, Canada

Scientific Session • Relaxometry - Methods & Corrections

Tuesday 2 June 2015

Room 701 B

16:00 - 18:00


Sean C. L. Deoni, Ph.D., Marcel Warntjes, Ph.D.

16:00 0439.   
Simultaneous Quantitative Mapping of T1, T2*, and Magnetic Susceptibility with Multi-Echo MP2RAGE at 7 T
Riccardo Metere1, Harald E. Möller1, Gunnar Krüger2,3, Tobias Kober2,3, and Andreas Schäfer1
1Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany, 2Siemens ACIT – CHUV Radiology, Siemens Healthcare IM BM PI & Department of Radiology CHUV, Lausanne, Switzerland, 3LTS5, École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland

Knowledge of the relaxation times is essential for understanding the biophysical mechanisms underlying image contrast, and can be related to tissue composition. Quantitative experiments are essential for cross-study comparisons, but often require longer acquisition times. Using the prototype Multi-Echo (ME) MP2RAGE pulse sequence, we show the possibility of obtaining reliable T1, T2*, and magnetic susceptibility maps that are inherently co-registered. The acquisition time required for such experiments is comparable to the time required for T1 map acquisition with MP2RAGE. This result is obtained as a trade-off between resolution and the parameters of the multi-echo readouts.

16:12 0440.   Fast T1 Mapping Using Slice-Shuffled Simultaneous Multi-Slice Inversion Recovery EPI
Hua Wu1, Robert F Dougherty1, Adam B Kerr2, Kangrong Zhu2, Matthew J Middione3, and Aviv Mezer4
1Center for Cognitive and Neurobiological Imaging, Stanford University, Stanford, CA, United States, 2Electrical Engineering, Stanford University, Stanford, CA, United States, 3Applied Sciences Laboratory West, GE Healthcare, Menlo Park, CA, United States, 4Psychology, Stanford University, Stanford, CA, United States

T1 mapping provides useful information for quantitative MR. Gold-standard method using inversion recovery spin echo sequence is robust but slow. EPI readout together with simultaneous multi-slice excitation can greatly accelerate the acquisition, and using a slice-shuffled method can accelerate the sampling of the T1 recovery. Here we test and validate fast T1 mapping methods using a combination of simultaneous multi-slice excitation to accelerate the cross-slice acquisition, in-plane acceleration to shorten the readout, and slice-shuffling to achieve rapid T1 recovery sampling.

16:24 0441.   T2-Snapshots imaging with simultaneous multislice TESS acquisition
Orso Pusterla1, Francesco Santini1, Rahel Heule1, and Oliver Bieri1
1Radiological Physics, Department of Radiology, University of Basel Hospital, Basel, Switzerland

Single slice triple echo steady-state (TESS) imaging has demonstrated potential for robust highly B1-insensitive T2 relaxometry in the human brain. In this work, a Hadamard-encoding excitation scheme is investigated for fast simultaneous multislice TESS acquisitions. The achieved improvement in the signal-to-noise ratio was invested into a reduction of the overall scan time per slice yielding about 10 sec / slice. As a result, simultaneous multislice TESS imaging is of high interest for accurate T2 quantification in the clinical routine offering full brain coverage within clinically acceptable scan times.

16:36 0442.   Artifact-free T2* mapping without post hoc corrections
Pippa Storey1, Yvonne W. Lui1, and Dmitry S. Novikov1
1Radiology Department, New York University School of Medicine, New York, NY, United States

Macroscopic susceptibility differences are a frequent source of artifacts in T2* mapping, and post hoc correction methods are often cumbersome or inaccurate. We demonstrate that Fourier encoding is largely insensitive to macroscopic phase gradients up to a threshold determined by the Nyquist limit. Accurate T2* maps can thus be obtained without post hoc correction simply by acquiring data in 3D mode and truncating the echo train where the local phase gradient exceeds the threshold. This approach produces uniform T2* maps in phantoms despite the presence of a ferromagnetic object, and artifact-free T2* maps of the brain even near the sinuses.

16:48 0443.   Fast dynamic measurements of T1 relaxation times: influence and correction of T2* effects
Olaf Dietrich1, Maximilian Freiermuth1, Linus Willerding2, Michael Peller1, and Maximilian F Reiser1
1Josef Lissner Laboratory for Biomedical Imaging, Institute for Clinical Radiology, LMU Ludwig Maximilian University of Munich, Munich, Germany,2Department of Internal Medicine III, LMU Ludwig Maximilian University of Munich, Munich, Germany

The purpose of this study was to analyze and correct the influence of contrast-agent induced T2* (transverse) relaxation effects on the accuracy of fast dynamic 3D gradient-echo T1 measurements with a combined variable-flip-angle baseline measurement and subsequent single-flip-angle measurements during the dynamic phase. Previously proposed techniques for T1 quantitation neglect the influence of transverse relaxation effects on the measured signal during the dynamic phase and result in systematically increased T1 values. We demonstrate that the influence of transverse relaxation can be expressed as a function of T1 leading to an implicit expression for T1 that can be solved numerically.

17:00 0444.   
Capital Greek DeltaB0 Correction for Myelin Water Fraction Imaging Based on Multi-Slice MGRE Acquisitions
Eva Alonso Ortiz1, Ives R. Levesque2,3, and G. Bruce Pike4
1McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada, 2Medical Physics Unit, Department of Oncology, McGill University, Montreal, Quebec, Canada, 3Research Institute of the McGill University Health Centre, McGill University, Montreal, Quebec, Canada, 4Department of Radiology and Hotchkiss Brain Institute, University of Calgary, Alberta, Canada

Myelin water fraction (MWF) imaging based on multi-component (MC) T2 analysis can be used to quantitatively assess myelin. Recently, multi-gradient echo (MGRE) sequences have been proposed for MCT2*-based MWF mapping, due to their fast, whole-brain or multi-slice imaging capability, low specific absorption rate (SAR) and short echo spacing (ES). An important caveat for MGRE imaging is the significant signal loss caused by magnetic field inhomogeneities (ΔB0), making proper ΔB0 correction necessary in order to avoid inaccuracies in MWF maps. We propose a ΔB0 correction method based on partial fits to MGRE data obtained for the MWF-mapping itself.

17:12 0445.   Encoding with Radiofrequency Spoiling, Equilibrium States and Inverse Problem for Parametric Mapping
Ludovic de Rochefort1
1IR4M (Imagerie par Résonance Magnétique Médicale et Multi-modalités), Univ. Paris-Sud, CNRS, UMR8081, Orsay, France
Fast steady-state sequences often use RF spoiling to modulate contrast in MRI. The dynamic equilibrium depends on many parameters such as flip angle, longitudinal and transverse relaxations, and diffusion. Here it is shown that, extending the configuration state formalism, RF spoiling can be described as an encoding direction and equilibrium can be efficiently calculated with relaxation and diffusion. The inverse problem can then be solved to reconstruct associated parametric maps. Theory, practical implementation as well as proof-of-concept experiments on a clinical MRI system are given introducing new methods for efficient parametric mapping.

17:24 0446.   The effect of macroscopic field gradients on the simultaneous estimation of reversible and irreversible transverse relaxation rates
Mukund Balasubramanian1,2 and Robert V. Mulkern1,2
1Department of Radiology, Boston Children's Hospital, Boston, MA, United States, 2Harvard Medical School, Boston, MA, United States

Pulse sequences such as GESFIDE and GESSE have enabled the simultaneous measurement of irreversible and reversible transverse relaxation rates. Reversible relaxation is influenced by both macroscopic and mesoscopic field inhomogeneities, with considerable interest in separating effects of the former from the latter. Here, we applied linear macroscopic gradients during GESSE scans of the knee, where significant mesoscopic field variations are induced by bone trabeculae. We find that the sinc-term corrections that have been proposed are not warranted and that the recently-observed Gaussian behavior of GESSE time-domain signals provides a better avenue for factoring out the effects of macroscopic field variations.

17:36 0447.   
Simultaneous group-wise rigid registration and Maximum Likelihood T1 estimation for T1 mapping
Gabriel Ramos-Llordén1, Arnold J. den Dekker1,2, Gwendolyn Van Steenkiste1, Johan Van Audekerke3, Marleen Verhoye3, and Jan Sijbers1
1iMinds-Vision Lab, University of Antwerp, Antwerp, Belgium, 2Delft Center for Systems and Control, Delft University of Technology, Delft, Netherlands, 3Bio-Imaging Lab, University of Antwerp, Antwerp, Belgium

In T1 mapping, to prevent motion artifacts, alignment of the acquired T1 weighted images is required. Commonly, image registration is accomplished prior to T1 map estimation. However, this two-step approach introduces bias in the T1 estimation due to inaccurate motion estimation and image interpolation. We propose a simultaneous group-wise rigid registration and T1 estimation method using a Maximum Likelihood (ML) approach for brain T1 mapping, thereby constructing a unified framework and circumventing the problems of the conventional two-step approach. Results with synthetic and real data demonstrate that the proposed method outperforms the conventional two-step approach.

17:48 0448.   
Field Probes with In-Situ Controllable Thermal Relaxation Times
David O. Brunner1, Simon Gross1, Jennifer Nussbaum1, Benjamin E. Dietrich1, Christoph Barmet1,2, and Klaas P. Pruessmann1
1Institute for Biomedical Engineering, University and ETH Zurich, Zurich, Switzerland, 2Skope Magnetic Resonance Technologies LLC, Zurich, Switzerland

For optimal performance, the T 1 and T 2 relaxation times of NMR based field sensors should be appropriately chosen compared to the repetition time of their re-excitation in the envisioned application. Too fast relaxations would prevent acquiring long read-outs, however too fast re-excitation would generate confounded field values due to occurring spurious echoes in the sensor. To date this is achieved by paramagnetic dopants added when manufacturing the probes but in this work we present an approach based on electrochemical reactions allowing tuning the relaxation time of an existing sensor between reversibly 3 ms and 60 ms.