Original Research T 1 Mapping Using Variable Flip Angle SPGR Data With Flip Angle Correction Gilad Liberman, MS, 1,2 Yoram Louzoun, PhD, 2,3 and Dafna Ben Bashat, PhD 1,4 * Purpose: To improve the calculation of T 1 relaxation time from a set of variable flip-angle (FA) spoiled gradient recalled echo images. Materials and Methods: The proposed method: (a) uses a uniform weighting of all FAs, (b) takes into account global inaccuracies in the generation of the prescribed FAs by estimating the actual FAs, and (c) incorporates data-driven local B 1 inhomogeneity corrections. The method was validated and its accuracy tested using simu- lated data, phantom, and in vivo experiments. Results were compared with existing analysis methods and to inversion recovery (IR). Consistency was assessed by means of repeated scans of two subjects. Reference values were obtained from eight healthy subjects from various brain regions and compared with literature values. Results: The method accurately and consistently esti- mated T 1 values in all cases. The method was more robust, in comparison with the standard method, to the choice of FA set; to inaccuracies in generation of the pre- scribed FAs (in simulated data, T 1 estimation error was 12.1 ms versus 235.5 ms); demonstrated greater consis- tency (in vivo study showed interscan T 1 difference of 80 ms versus 356 ms); and achieved a better agreement with IR on phantom (median absolute difference of 123.8 ms versus 790 ms). Reference T 1 values were 883/801 ms for female/male in white matter and 1501/1349 ms in gray matter, within the range previously reported. Conclusion: The proposed method overcomes some inaccuracies in FA production, providing more accurate estimation of T 1 values compared with standard methods, and is applicable for currently available data. Key Words: relaxometry; T 1 mapping; FA correction, SPGR J. Magn. Reson. Imaging 2013;00:000–000. V C 2013 Wiley Periodicals, Inc. ACCURATE ESTIMATION OF the longitudinal relaxation time, T 1 , of brain tissues has multiple clinical applica- tions, and plays a significant role in the shift toward quantitative analysis of MRI signals. In clinical research of several neurological disorders, such as multiple scle- rosis and Parkinson’s disease (1,2), T 1 values improve the delineation of specific brain structures and the differentiation between patients and healthy subjects. In addition, accurate T 1 values are necessary for the extraction of pharmacokinetic parameters, from dynamic contrast enhanced (DCE) data (3). Additionally, knowledge of the T 1 value leads to better approximation of the T 2 value in certain methods, such as driven equi- librium single pulse observation of T 2 (DESPOT2) (4). To this end, various MRI sequences and analysis methods have been proposed to calculate the T 1 value (4–11). One of these methods is the variable flip-angle (VFA) spoiled gradient recalled echo (SPGR) method, that provides high spatial resolution with relatively short acquisition times (1,4), and is commonly used in basic and clinical research. Significant efforts have been devoted to developing a fast and highly accurate analysis (12) of the VFA-SPGR data and to finding the best set of flip angles (FAs) that maximize accuracy and efficiency (in terms of T 1 value to noise ratio per acquisition time unit), for a specific T 1 value or a range of values (4,13,14). There is currently no broadly accepted FA set, and in many institutions, a varied set of angles is used. In addition, most current methods weight each FA dataset by a different factor that depends on the FA, resulting in a nonuniform weighting of the datasets. Thus, the calculated maps are affected by the choice of the FA set (13). Another limitation of most current analysis methods is that they base their calculation of the T 1 values on the prescribed FA as reported by the MRI system. Yet, the actual FA might deviate from the one reported by the system, both globally, which may occur due to inaccuracy in calibration, and locally, due to inhomo- geneity in the radiofrequency (RF) transmit field B 1 . Several sequences have been proposed to determine 1 The Functional Brain Center, The Wohl Institute for Advanced Imag- ing, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel. 2 Gonda Multidisciplinary Brain Research Center, Bar-Ilan University, Ramat Gan, Israel. 3 Department of Mathematics, Bar-Ilan University, Ramat Gan, Israel. 4 Sacklar Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel. Additional Supporting information may be found in the online version of this article. Contract grant sponsor: James S. McDonnell Foundation; Contract grant number: 220020176. *Address reprint requests to: D.B.B., The Wohl Institute for Advanced Imaging, Brain Imaging Center, Tel Aviv Sourasky Medical Center, 6 Weizmann Street, Tel Aviv, 64239, Israel. E-mail: dafnab@tlvmc.gov.il Received February 28, 2013; Accepted July 14, 2013. DOI 10.1002/jmri.24373 View this article online at wileyonlinelibrary.com. JOURNAL OF MAGNETIC RESONANCE IMAGING 00:00–00 (2013) CME V C 2013 Wiley Periodicals, Inc. 1