Influence of the pressure field distribution in transcranial ultrasonic neurostimulation
Youliana Younan, Thomas Deffieux, Benoit Larrat, Mathias Fink, Mickael Tanter, Jean-Francois Aubry
Medical Physics 2013, 40, 082902 · 10.1118/1.4812423
Abstract
Purpose Low-intensity focused ultrasound has been shown to stimulate the brain noninvasively and without noticeable tissue damage. Such a noninvasive and localized neurostimulation is expected to have a major impact in neuroscience in the coming years. This emerging field will require many animal experiments to fully understand the link between ultrasound and stimulation. The primary goal of this paper is to investigate transcranial ultrasonic neurostimulation at low frequency (320 kHz) on anesthetized rats for different acoustic pressures and estimate the in situ pressure field distribution and the corresponding motor threshold, if any. The corresponding acoustic pressure distribution inside the brain, which cannot be measured in vivo, is investigated based on numerical simulations of the ultrasound propagation inside the head cavity, reproducing at best the experiments conducted in the first part, both in terms of transducer and head geometry and in terms of acoustic parameters. Methods In this study, 37 ultrasonic neurostimulation sessions were achieved in rats (N=8) using a 320 kHz transducer. The corresponding beam profile in the entire head was simulated in order to investigate the in situ pressure and intensity level as well as the spatial pressure distribution, thanks to a rat microcomputed tomography scan (CT)-based 3D finite differences time domain solver. Results Ultrasound pulse evoked a motor response in more than 60% of the experimental sessions. In those sessions, the stimulation was always present, repeatable with a pressure threshold under which no motor response occurred. This average acoustic pressure threshold was found to be 0.68±0.1 MPa (corresponding mechanical index, MI=1.2 and spatial peak, pulse averaged intensity, Isppa=7.5 W cm(-2)), as calibrated in free water. A slight variation was observed between deep anesthesia stage (0.77±0.04 MPa) and light anesthesia stage (0.61±0.03 MPa), assessed from the pedal reflex. Several kinds of motor responses were observed: movements of the tail, the hind legs, the forelimbs, the eye, and even a single whisker were induced separately. Numerical simulations of an equivalent experiment with identical acoustic parameters showed that the acoustic field was spread over the whole rat brain with the presence of several secondary pressure peaks. Due to reverberations, a 1.8-fold increase of the spatial peak, temporal peak acoustic pressure (Psptp) (±0.4 standard deviation), a 3.6-fold increase (±1.8) for the spatial peak, temporal peak acoustic intensity (Isptp), and 2.3 for the spatial peak, pulse averaged acoustic intensity (Isppa), were found compared to simulations of the beam in free water. Applying such corrections due to reverberations on the experimental results would yield a higher estimation for the average acoustic pressure threshold for motor neurostimulation at 320 KHz at 1.2±0.3 MPa (MI=2.2±0.5 and Isppa=17.5±7.5 W cm(-2)). Conclusions Transcranial ultrasonic stimulation is pressure- and anesthesia-dependent in the rat model. Numerical simulations have shown that the acoustic pattern can be complex inside the rat head and that special care must be taken for small animal studies relating acoustic parameters to neurostimulation effects, especially at a low frequency.
Abstract via europepmc.
Exposures
Exposure 1: Transcranial sonication near lambda for motor stimulation (tail/hindleg/forelimb/eye/whisker responses)
Target: other — “around the locations lambda −1, 1, and 3 mm; these chosen locations did not correspond to the motor cortex area”
Device: Sonic Concepts · Sonic Concepts, Bothell, MA · H115 ✓
| Waveform | pulsed | |
|---|---|---|
| Fundamental frequency (kHz) | 320 | ✓✓✓ |
| Pulse duration (ms) | 0.23 | ✓✓✓ |
| Pulse repetition frequency (Hz) | 2,000 | ✓✓✓ |
| Duty cycle (%) | 50pulse duration × PRF gives 46%, which disagrees with the stated value | ✓✓✓ |
| Sonication duration (s) | 0.25 | ✓✓✓ |
| Free-field pressure (kPa) | 400, 1,000swept | ✓✓✓⚑ |
|---|---|---|
| Free-field Isppa (W/cm²) | 7.5 | ✓✓✓ |
| Free-field Ispta (W/cm²) | not reported | |
| In-situ estimate | simulationmean or range across subjects | |
| In-situ pressure (kPa) | 1,200 | ✓✓✓ |
| In-situ Isppa (W/cm²) | 17.5 | ✓✓✓ |
| In-situ Ispta (W/cm²) | not reported |
The ultrasound sequence was based on the protocol of Tufail et al., with a slightly higher center frequency (320 kHz instead of 300 kHz) and a longer total sonication duration of 250 ms instead of 100 ms, found to be more efficient. Ultrasound pulses were applied to the rat brain every 10 s, and for each series of measurements 30 different pressure amplitudes were tested in random order to build a motor-response-versus-pressure curve.
Consistency checks: intensity pressure inconsistent in situ.
Flags from extraction
randomised— Randomisation described applies to the order of pressure amplitudes tested within a session, not assignment of subjects to groups; there are no groups/arms in this single-arm dose-response design.exposures[0].target.terms— Target was a skull-surface location near lambda used because it gave reliable motor responses, explicitly not the motor cortex; no listed target term matches, so 'other' was used.exposures[0].free_field.pressure_kpa— Reported as the tested range (0.4-1 MPa); the mean motor threshold (0.68 MPa) and its in-situ correction are separately reported as results, not as the exposure list.