Decoding paradoxical BOLD responses to transcranial ultrasound stimulation with concurrent optoacoustic magnetic resonance imaging
Yi Chen, Zhenyue Chen, Hector Estrada, Irmak Gezginer, Hikari A. I. Yoshihara, Diana Kindler, Chunqi Qian, David C. Zhu, Shy Shoham, Daniel Razansky
Science Advances 2025 · 10.1126/sciadv.adz1309
Abstract
Focused transcranial ultrasound stimulation (TUS) can affect neural activity with high spatial precision, advancing noninvasive neuromodulation toward targeted treatment of brain disorders. Direct monitoring of TUS responses is crucial for ensuring optimal outcomes. Blood-oxygenation-level-dependent (BOLD) functional magnetic resonance imaging has primarily been used for studying TUS effects in both human and nonhuman primate brains. However, the physiology and mechanisms underlying BOLD remain largely unknown due to its highly convoluted nature. To address these limitations, we developed a hybrid system for concurrent optoacoustic and magnetic resonance imaging of TUS (OMRITUS) to comprehensively characterize the hemodynamic changes in murine brains. Our findings reveal paradoxical negative BOLD signals in the activated cortical regions, coupled with increased total hemoglobin levels simultaneously monitored with optoacoustic tomography. Multispectral optoacoustic readings further demonstrated a stronger increase in deoxygenated versus oxygenated hemoglobin, suggesting a potential molecular basis for the negative BOLD responses. OMRITUS enables the study of complex TUS-hemodynamic interactions, paving the way for precise neuromodulatory interventions.
Abstract via europepmc.
Exposures
Exposure 1: Holographic (multifocal) TUS of right RSC, concurrent fMRI and OAT
Target: retrosplenial cortex — “right RSC of the mouse brain”
Device: not reported · 512-element spherical array transducer (OMRITUS system) ✓
| Waveform | pulsed | |
|---|---|---|
| Fundamental frequency (kHz) | 3,000 | ✓✓✓ |
| Pulse duration (ms) | 100 | ✓✓✓⚑ |
| Pulse repetition frequency (Hz) | 4 | ✓✓✓ |
| Duty cycle (%) | 40pulse duration × PRF gives 40% | ✓✓✓ |
| Sonication duration (s) | 8, 20swept | ✓✓✓ |
| Free-field pressure (kPa) | not reported | |
|---|---|---|
| Free-field Isppa (W/cm²) | not reported | |
| Free-field Ispta (W/cm²) | not reported | |
| In-situ estimate | measurementsingle value | |
| In-situ pressure (kPa) | 300, 400, 520swept | ✓✓✓⚑ |
| In-situ Isppa (W/cm²) | not reported | |
| In-situ Ispta (W/cm²) | not reported |
A 100-ms ultrasound pulse at 3 MHz delivered at 4 Hz for 8 s (22 s interstimulation interval, 6 epochs) or 20 s (60 s interval, 5 epochs); pressure levels of 0, 0.3, 0.4 and 0.52 MPa were applied; three foci were sonicated simultaneously (holographic).
Flags from extraction
exposures[0].in_situ.pressure_kpa— 0.52 MPa is the focal pressure measured through an excised mouse skull; other levels (0.3, 0.4, text; 0.41 in Fig. 3 legend) are named in the protocol description only and the domain of those is not stated.n_subjects— 14 = 8 C57BL/6J + 6 nude mice 'used in this study'; not all were necessarily sonicated (group analyses use n = 6); postmortem controls were also run.sham_type— No sham arm; control sonications were delivered with the same parameters after euthanasia, plus 0 MPa rest-state scans.auditory_control— Auditory/visual cortex BOLD was inspected within the same fMRI data; no separate auditory control.exposures[0].device.family— Array maker is not named; DAQ credited to Falkenstein Mikrosysteme and described as customized.exposures[0].timing.pulse_duration_ms— Paper calls the 100 ms an 'emission duration'/'pulse duration'; with 4 Hz PRF and 40% duty cycle it is consistent as the pulse. Not resolved by arithmetic.