2711

Correction of LIFU Intensity Using MR Acoustic Radiation Force Imaging (ARFI): Improved Correlations with Treatment Outcomes
Abubakr Eldirdiri1, Linda Chang1, David Martin1, Donna Calu1, Eric Cunningham1, Segun Bernard1, Prathibha Meesala1, and Thomas Ernst1
1University of Maryland School of Medicine, Baltimore, MD, United States

Synopsis

Keywords: MR-Guided Focused Ultrasound, Focused Ultrasound

Motivation: The effectiveness of low-intensity focused ultrasound (LIFU) for neuromodulation may be compromised by inaccuracies in target localization and variable energy deposition caused by skull characteristics.

Goal(s): We aimed to improve the estimation of LIFU pressure using MR-acoustic radiation force imaging (ARFI) and test whether the adjusted LIFU intensity correlates with treatment outcomes in a behavioral economic task using sucrose pellets.

Approach: Eight female rats underwent LIFU targeting the nucleus accumbens (NAc) bilaterally with three different pressures. MR-ARFI confirmed the targeting and corrected the intensities during LIFU treatments.

Results: Improved correlations between behavioral outcomes and ARFI-adjusted LIFU intensity were observed.

Impact: Our preliminary findings demonstrate the beneficial effects for using MRI-ARFI not only to verify the FUS target location but also to refine LIFU intensities in neuromodulation procedures.

INTRODUCTION

Low intensity focused ultrasound (LIFU) is an emerging technique that can non-invasively modulate deep brain regions1-3, and may become a useful adjunct therapy for various neuropsychiatric disorders4-6. However, variations in skull density, thickness, and shape, and variations in transducer placement can induce focused ultrasound (FUS) aberrations and ultimately cause inaccurate localization of the target focal point and attenuation in energy deposition7.
MR acoustic radiation force imaging (ARFI)8-10 uses a bipolar gradient to detect microscopic movements caused by deposited ultrasound energy. ARFI is typically used to confirm the location of the focal point of a FUS beam, but the phase change induced also provides an estimate of the tissue displacement and ultimately the acoustic intensity delivered at the target.
Recent work showed that LIFU targeting the nucleus accumbens (NAc), the reward center of the brain, may reduce craving in individuals with substance use disorders5,6. This study aims to: 1) show how MR-ARFI can be used to calibrate the FUS intensity in LIFU experiments; 2) determine whether the adjusted FUS intensity correlates with LIFU treatment outcomes in a food reward rat model. We hypothesized that the MR-ARFI-adjusted FUS intensity will show better inverse correlation with the amount of food consumption after LIFU inhibition of bilateral NAc.

METHODS

Animal Model and Behavioral Studies
Eight Sprague-Dawley rats (8-week-old females) were trained to stable behavior and tested in a behavioral economic task using sucrose pellets. The task employed increasing fixed ratio requirements across bins within a session. Lever presses and pellets earned were recorded for each bin.
To suppress neural activity in the NAc, LIFU treatments targeting the NAc bilaterally were administered at pressures of 0.5, 1, or 1.5 MPa (calibrated in a phantom with a hydrophone, Figure 1A). The LIFU treatment scheme is illustrated in Figure 1B. On Day 1, 4 rats that were anesthetized for MRI-guided-ARFI-LIFU-treatments were each paired with another rat that had the same duration of anesthesia. On Day 2, one week later, those that had anesthesia only on Day 1 received the MR-guided-ARFI-LIFU, while those that received MRI-guided-ARFI-LIFU previously had anesthesia only. Food consumption measurements were conducted two hours after the MR-guided LIFU treatments or anesthesia only.
MR-Guided ARFI-LIFU
A 7T small animal MRI scanner (Bruker, Germany) equipped with an MR-compatible 1.5MHz MR-guided FUS system (Image Guided Therapy, Pessac, France) was used. This system has an 8-element annular array of FUS transducers and a single loop transmit-receive RF coil.
During the MR-ARFI-LIFU, T2-weighted axial and coronal MRI were performed to identify the target regions (Figure 2A-C). Next, MR-ARFI was performed at fixed power to validate the acoustic coupling and focal spot location in the brain (Figure 2D,E). The MR-ARFI sequence diagram and parameters, at fixed FUS power, are shown (Figure 3). Two sets of MRI-ARFI phase images were acquired, one with and one without sonication (reference phase image) and were subtracted to obtain the phase-encoded displacement.
Since the ARFI-FUS power was constant across animals, variations in the ARFI-derived displacement reflect experimental FUS-intensity variations at the focal point. The ARFI-derived displacement values were normalized across the 8 rats to yield $$$D_i$$$ (i=1:8), and used to calculate an adjusted LIFU intensity $$${L^a}_i$$$ for each animal:
$${L^a}_i=D_i{P^t}_i [1]$$
where $$${P^t}_i$$$ is the LIFU pressure set during treatment of the animal (0.5, 1, or 1.5MPa).

RESULTS

Across the three conditions, significant differences were observed with lever presses, pellets acquired, and maximum consumption (repeated-measure ANOVA-p<0.05). Post-hoc comparisons showed significantly lesser food consumption post-LIFU treatments (p<0.05), but not after anesthesia-only, relative to the baseline (Figure 4).
Although ARFI was performed at the same FUS intensity, the measured displacement varied between 0.80 and 2.7 μm. No apparent correlation was found between the administered LIFU dosage and both treatment outcomes (Figure 5A,B). However, when using the adjusted LIFU intensity $$${L^a}_i$$$ (Eq.[1]), correlations emerged between the ARFI-estimated LIFU intensity and treatment outcomes (Figure 5C,D).

DISCUSSION

LIFU intensity measured with ARFI deviated substantially from intended values. We found improved correlations between behavioral outcome measures with the corrected acoustic pressure determined using ARFI. These preliminary findings suggest NAc LIFU reduces food motivation, while residual effects of anesthesia at 2 hours post treatment cannot be ruled out. Limitations include: 1) lack of a control group receiving sham LIFU treatments or LIFU in an active control brain region unrelated to reward-seeking behavior; 2) Further work will determine the recovery period required to normalize the behavior after ARFI-LIFU treatments.

CONCLUSION

This study highlights the benefits of using ARFI not only to verify the FUS target location, but also to refine the accuracy of LIFU intensities in neuromodulation procedures.

Acknowledgements

This work was supported by grants from the Focus Ultrasound Foundation Grant (to L.C.) and the National Institute on Drug Abuse (DP1DA053719 to L.C.)

References

  1. Arulpragasam AR, van't Wout-Frank M, Barredo J, et al. Low intensity focused ultrasound for non-invasive and reversible deep brain neuromodulation—a paradigm shift in psychiatric research. Front Psychol. 2022;13:825802.
  2. Folloni D, Verhagen L, Mars RB, et al. Manipulation of subcortical and deep cortical activity in the primate brain using transcranial focused ultrasound stimulation. Neuron. 2019;101(6):1109-1116.
  3. Yüksel MM, Sun S, Latchoumane C, et al. Low-intensity focused ultrasound neuromodulation for stroke recovery: A novel deep brain stimulation approach for neurorehabilitation?. IEEE OJEMB. 2023.
  4. Krishna V, Sammatino F, Rezai A. A Review of the Current Therapies, Challenges, and Future Directions of Transcranial Focused Ultrasound Technology: Advances in Diagnosis and Treatment. JAMA Neurol. 2018;75(2):246-254.
  5. Mahoney JJ, Thompson-Lake DG, Ranjan M, et al. Low-Intensity Focused Ultrasound Targeting the Bilateral Nucleus Accumbens as a Potential Treatment for Substance Use Disorder: A First-in-Human Report. Biol Psychiatry. 2023;94(11):e41-43.
  6. Mahoney J, Haut MW, Carpenter J, et al. Low-intensity focused ultrasound targeting the nucleus accumbens as a potential treatment for substance use disorder: safety and feasibility clinical trial. Front Psychol. 2023;14:1211566.
  7. Kyriakou A, Neufeld E, Werner B, et al. A review of numerical and experimental compensation techniques for skull-induced phase aberrations in transcranial focused ultrasound. Int J Hyperth. 2014;30(1):36-46.
  8. Larrat B, Pernot M, Aubry JF, et al. MR-guided transcranial brain HIFU in small animal models. Phys Med Biol. 2009;55(2):365-388.
  9. Chen J, Watkins R, Pauly KB. Optimization of encoding gradients for MR‐ARFI. Magn Reason Med. 2010;63(4):1050-1058.
  10. Gaur P, Casey KM, Kubanek J, et al. Histologic safety of transcranial focused ultrasound neuromodulation and magnetic resonance acoustic radiation force imaging in rhesus macaques and sheep. Brain Stimul. 2020;13(3):804-814.

Figures

Figure 1: (A) displays the results of FUS calibration. This calibration was performed on an ultrasound gel phantom using a hydrophone and an oscilloscope to estimate the pressure at the focal point across various input levels of electric power. (B) shows the LIFU treatment scheme used to sonicate the nucleus accumbens (NAc). The FUS treatment comprised 180 bursts delivered over a period of 15 minutes. The pulse repetition frequency, pulse duration, burst duration, and duty cycle were denoted by PRF, PD, BD, and DC, respectively.

Figure 2: (A, B, and C) depict the planning of the low-intensity focused ultrasound (LIFU) treatment on T2-weighted images targeting the NAc in one side of the brain. These images were acquired with the following parameters: repetition time (TR) = 3000 ms, echo time (TE) = 36 ms, slice thickness = 1 mm, and a total of 30 slices. Phase encoded displacement fields (in radians) are illustrated in (D and E), obtained through acoustic radiation force imaging (ARFI) acquisition.

Figure 3: The sequence diagram of the ARFI process. The ARFI procedure was conducted with TE = 25 ms, TR = 1000 ms, field of view (FOV) = 50 × 50 mm², a matrix size of 64 × 64, and a slice thickness of 2 mm. The duration of each lobe in the bipolar gradients was 4 ms, with a maximum strength of 135 mT/m. The ultrasound power was consistently set to 16 W (equivalent to 50% of the maximum power) for all ARFI experiments. Additionally, the ultrasound pulse duration was fixed at 10 ms to encompass the two central lobes within the bipolar gradient.

Figure 4: The number of lever presses and sugar pellets, and maximum consumption at baseline, post-anesthesia-only, and post-LIFU treatment. The posthoc p-values reported on the bargraphs are from two-sample t-tests. The overall p-values on both the bargraphs and the line-graphs are from repeated-measure ANOVA. One animal died during the MRI-ARFI-LIFU (only 7 animals had the post-LIFU behavioral measures).

Figure 5: A and B show no apparent correlation between the raw LIFU dosage and the change in substance seeking behaviour, (postLIFU-baseline)/baseline. Estimating the FUS intensity using ARFI, reveals improved correlations between the adjusted FUS intensity and the change in substance seeking behavior as shown in C and D.

Proc. Intl. Soc. Mag. Reson. Med. 32 (2024)
2711
DOI: https://doi.org/10.58530/2024/2711