hide
Free keywords:
-
Abstract:
Introduction: The ’’dual refocusing echo acquisition mode‘‘ (DREAM) sequence is a fast method for B0 and B1 field mapping in 2D [1], and 3D [2]. DREAM also provides a Tx/Rx-phase mapping. We implemented a dynamic DREAM sequence that provides all these information, and evaluated a dynamic scan under free breathing, and a scan where the head position in the coil was altered. With up to 1 s temporal resolution breathing induced B0 could be detected; also motion induced B0, B1 and Tx/Rx-phase changes could be detected. This information can in the future be used to assist impedance-based, optical or radar-based motion tracking systems. Methods: A centric-reordered DREAM sequence [1] (matrix: 64 9 64, FoV = 220 9 220 9 8 mm 3 , FA STE1 = FA STE2 = 55°, FA = 15°, TE FID = 3.6 ms, TE STE = 2.6 ms, TR = 5.6 ms, TA = 387 ms) was realized in Pulseq and used with a Pulseq interpreter sequence. At submission time, a first healthy volunteer was scanned under approval of our local ethics committee on a 3 T whole-body MRI system (MAGNETOM Prisma, Siemens Healthcare, Erlangen, Germany) and a 20 channel receive coil. The DREAM provides for every image a B0 and B1 map, and a TxRx-phase map calculated by the formulas [1]: UB0 ¼ angle FID STEð Þ ð4Þ rB1 ¼ tan1 p 2 STEj j=jFIDjð Þð Þ=FA STE1 ð5Þ Utx=rx ¼ angle FID STEð Þ ð6Þ In the dynamic free-breathing measurement, the DREAM was repeated 20 times with a recovery time in-between sequences of 2 s leading to a temporal resolution of 2.4 s. Here a pre-experiment showed that 1 s delay yields unaltered field maps (data not shown). In the positioning measurement, the DREAM was scanned for four specific positions of the head in the coil. The FoV was fixed with regard to the brain using the head-scout auto-align functionality. This allows calculating exact difference maps of the different positioned images. Results: The dynamic free-breathing measurements showed that B1 and TxRx-phase seems to be stable with breathing, but B0 shows the typical breathing state induced deviations (Fig. 1). Figure 2 shows that the dynamic B0 shift and the FFT which peaks at the breathing frequency. The positioning experiment revealed that the B0 and B1 field maps show non-linear alterations upon translation of the head (Fig. 3), while the TxRx-phase (Fig. 4) seems to translate linearly with the movement (in the head frame). I.e. the TxRx-phase seems to be almost stationary in the lab frame. Discussion: DREAM MRI can be used to dynamically map B0, B1, and the TxRx-phase with temporal resolution up to 1 frame-per- second (FPS). This makes possible to analyze the influence of physiological processes on all these maps, as well as the influence of positioning or patient movement within the scanner and the shim- coils and the Tx- and Rx-coils. 3D implementations with TA of 4.2 s [2] should be possible with around 5 s per frame, thus 0.2 FPS. This dynamic field and phase map information is useful for assisting of motion tracking systems and corresponding dynamic field map estimation. Conclusion Dynamic DREAM MRI mapping of B0, B1, and the TxRx-phase is possible with 1 FPS.
