MR acoustic radiation force imaging (MR-ARFI) provides a method to visualize the focus of high-intensity focused ultrasound (HIFU) based on tissue stiffness¹. When ultrasound is absorbed by tissue, a force is exerted along the beam direction and can cause tissue displacement of a few micrometers at the focal spot when using HIFU transducers. By inserting a pair of motion-encoding gradients (MEG) into MR pulse sequences, with one lobe synchronized with a HIFU pulse, spins at the ultrasound focus accrue an additional phase that can be observed in MR phase images. MR-ARFI is mostly performed with spin-echo or spoiled GRE sequences², though fluctuating-equilibrium balanced steady-state free precession sequences have also been proposed³. In this work, MR-ARFI using transition band bSSFP is presented with a scan time of only a few seconds. This new technique improves the sensitivity of MR-ARFI measurements over spoiled sequences, by generating a larger phase change for a given displacement. Transition band bSSFP signal is highly sensitive to the amount of phase that spins accrue in the rotating frame during each TR. This phase β_off is usually caused by the difference between the spin resonance frequency and the scanner operating frequency. The highly sensitive transition band bSSFP signal has been used for fMRI⁴ and thermometry⁵. For bSSFP-ARFI, a pair of MEG are added to traditional bSSFP sequences (Fig. 1). By synchronizing a HIFU pulse with one lobe of the MEG, an additional phase β_MEG is generated by the radiation force displacement. We define the sensitivity of MR-ARFI pulse sequences as the derivative of the image phase ϕ with respect to the motion-encoded phase β_MEG. The sensitivity is always 1 with conventional pulse sequences. However, for the bSSFP-ARFI sequence in Fig.1, it can be shown that the image phase is: $$$\phi=\arctan(\frac{E_2\sin(\beta_{off}+\beta_{MEG})}{1-E_2\cos(\beta_{off}+\beta_{MEG})})+\frac{TE}{TR}\beta_{off}+\beta_{MEG}$$$ (Eq. 1), where E₂=exp(-TR/T₂). The ARFI sensitivity $$$(\frac{d\phi}{d\beta_{MEG}}\Bigr|_{\beta_{MEG}=0})$$$ is plotted as a function of β_off in Fig. 2. The highest sensitivity occurs on resonance and equals T₂/TR+1/2. Since T₂ is usually much longer than TR for bSSFP sequences, MR-ARFI sensitivity is amplified near resonance.A homemade gel phantom was used for this study. All MR images were acquired on a 3T GE MR750 scanner (GE Healthcare, Milwaukee, WI) with a single-channel local coil. Pulse sequences were developed and implemented on the scanner using RTHawk (HeartVista, Inc., Menlo Park). 2D images were collected with slice thickness=3 mm, FOV=12.8 cm, matrix=128x128. HIFU was generated by a multi-element phased-array transducer (Exablate 2000, Insightec, Haifa, Israel). The pulse duration was 1.45 ms and the acoustic power was 64.8 W. The HIFU beam was parallel to the MEG and perpendicular to the imaging slice. The bSSFP-ARFI images were acquired with FA=2°, BW=25 kHz, TE=10.5 ms, TR=14.7 ms. Each lobe of the MEG was 3 ms long, and the amplitude and duration of the plateau were 5 G/cm and 1.25 ms. Scanner frequency and shimming gradients were adjusted to ensure the expected focal spot position on resonance. Two images were acquired. HIFU was off for the first image as a baseline and on for the second one. The two images were acquired in 3.8 s. To study the effect of possible small temperature rise due to HIFU pulses on the image phase, we repeated the measurement with the same bSSFP sequence with MEG removed. To compare the enhanced sensitivity of the bSSFP-ARFI technique with traditional MR-ARFI, we kept MEG and HIFU the same as the bSSFP-ARFI experiment and performed MR-ARFI using a standard 2DFT spGRE-ARFI sequence with FA=40°, BW=16.125 kHz, TE=11.3 ms and TR=500 ms. Images were acquired with opposite MEG polarities (with separate baseline images). The total acquisition time for spGRE-ARFI was 4 min.Fig. 3a-c shows the phase contrast calculated from bSSFP-ARFI images, bSSFP images and the spGRE-ARFI images. The average phases in a 2 by 2 pixel ROI at the focal spot are 0.38, 0.10 and 0.052 respectively. Removing phase change caused by temperature change results in bSSFP-ARFI phase contrast of (0.38-0.10)=0.28. The phase contrast from the spGRE images equals β_MEG. Therefore the contrast of the bSSFP-ARFI sequence is amplified by (0.28/0.052)=5.4 times.We have developed a novel MR-ARFI technique that makes use of the sharp phase transition of bSSFP signals to amplify the motion-encoded phase. With its high speed and high sensitivity, the bSSFP ARFI technique could be useful in confirming/calibrating the HIFU focal spot before thermal ablation treatment.