Burst Excitation for Quantitative Diffusion Imaging With Multiple b-Values C.A. Wheeler-Kingshott, 1 * D.L. Thomas, 2 M.F. Lythgoe, 3 D. Guilfoyle, 4 S.R. Williams, 5 and S.J. Doran 6 A quantitative imaging sequence has been developed to exploit the intrinsic sensitivity of Burst NMR data to molecular diffu- sion. In the scan time of a single spin echo experiment, it is possible to acquire many images of the same slice, with a different T 2 and diffusion weighting. Under favorable condi- tions, it is possible to obtain both the diffusion coefficient and T 2 from the same experiment; or, by correcting for T 2 relaxation using a control image, more precise diffusion coefficients may be measured. The quantitative values in rat brain are in agree- ment with those from conventional experiments. The major gains of this method are the potentially reduced scan time, the higher number of acquired images corresponding to different diffusion weightings, the reduced sensitivity to inter-scan mo- tion artifact and to local variations in magnetic susceptibility, and an automatic co-registration between T 2 and diffusion im- ages. Problems with the sequence include a lower signal-to- noise ratio than is achievable with diffusion-weighted spin-echo imaging, the limitation of measuring only in-plane components of diffusion and, at present, single-slice acquisition. Magn Re- son Med 44:737–745, 2000. © 2000 Wiley-Liss, Inc. Key words: MRI; Burst; diffusion imaging; multiple b-factor NMR is well known to be sensitive to random molecular displacements, allowing one to measure the self-diffusion coefficients D of fluids and “soft solids” over a wide range, from approximately 10 -4 –10 -15 m 2 s -1 (1). The value of D reflects the mobility of the molecules in their microenvi- ronment (2) and can reveal information about the micro- scopic structure of the system under investigation. The high sensitivity to molecular displacements of the order of microns (three orders of magnitude less than the spatial resolution of a typical clinical MRI scan) makes DWI very sensitive to biophysical changes related to pathologies, even at the very early stages of their development. The clinical potential of diffusion weighted imaging (DWI) is most likely to be realized in the study of acute stroke, both for early detection of cerebral infarction and for monitoring subsequent ischemic damage. Changes in the value of the apparent diffusion coefficient (ADC) occur in affected brain regions within minutes of the cessation of blood flow, whereas relaxation time changes are not de- tected for some hours (3,4). In animal models, measure- ment of the ADC has been used to evaluate infarct devel- opment and to investigate the efficacy of experimental therapies (5– 8). The use of the technique clinically is now becoming routine for the diagnosis of stroke (9,10) and many other conditions (11–16). Diffusion imaging is also of considerable interest from a technical standpoint. Achieving measurements that are both accurate and rapid represents a considerable chal- lenge. The original Stejskal-Tanner pulse field gradients (PFG) method (17), usually involves the acquisition of one “unweighted” spin-echo image and one or more diffusion- weighted images. This is relatively slow, and: (a) Sample motion between the acquisition of separate images can lead to misregistration errors during pro- cessing and the generation of incorrect ADC values. (b) Motion between successive phase-encode steps of each individual image leads to an incorrect phase relation between the different lines of k-space data and thus to characteristic ghosting artifacts. This is a general feature of multi-shot sequences and is not specific to diffusion measurements (18). (c) Coherent motion during a single phase-encoding step can cause large additional phase errors, since a diffusion experiment is ipso facto very sensitive to motion. One solution, which works well for intra-cranial scans, is to use a PFG sequence with an echo-planar imaging (EPI) readout (19). This eliminates the artifacts due to (b) and drastically reduces the other contributions. In addition, the method is fast and is thus more economical, better tolerated by patients, and allows one to follow an evolving pathology with good temporal resolution. However, EPI has a number of well known problems. Particularly rele- vant are: 1) the requirement of a relatively long T 2 for the sample being measured, as time is needed for both the diffusion sensitization and readout gradients; and 2) “sus- ceptibility” artifacts in regions of inhomogeneous mag- netic field. The latter are often intrinsic to the sample being studied and the effect becomes worse with increas- ing field strength. Alternative fast techniques have been implemented, such as a PFG version of the turbo spin-echo sequence (20), a magnetization-prepared fast low angle shot (FLASH) technique (21), and line scan methods (22,23). Here, we examine the use of the Burst excitation scheme (24), which produces a signal that is inherently diffusion- and T 2 -weighted. A previous study by Doran and De ´corps (25) has shown that in “spectroscopic” mode, one can use it to extract the bulk diffusion and T 2 values of a sample in a single-shot (40 msec). With the introduction by Zha and Lowe (26) of a phase cycling scheme that increases the 1 NMR Unit, Department of Clinical Neurology, Institute of Neurology, UCL, London, UK. 2 Department of Medical Physics, UCL, London, UK. 3 Institute of Child Health, UCL, London, UK. 4 Nathan S. Kline Institute for Psychiatric Research, NY. 5 Imaging Science and Biomedical Engineering, University of Manchester, UK. 6 Physics Department, University of Surrey, UK. *Correspondence to: Dr. C. Wheeler-Kingshott, NMR Unit, Dept. of Clinical Neurology, Institute of Neurology, University College London, Queen Square, London WC1N 3BG, UK. E-mail: c.wheeler-kingshott@ion.ucl.ac.uk Received 1 November 1999; revised 16 June 2000; accepted 19 June 2000. Magnetic Resonance in Medicine 44:737–745 (2000) © 2000 Wiley-Liss, Inc. 737