Controlled delivery of ultrasound through the head for effective and safe therapies of the brain
Tom Riis, Matthew Wilson, Jan Kubanek
2022 · 10.1101/2022.12.16.520788
Abstract
Transcranial focused ultrasound provides noninvasive and reversible approaches for precise and personalized manipulations of brain circuits, with the potential to transform our understanding of brain function and treatments of brain dysfunction. However, the effectiveness and safety of these approaches have been limited by the human head, which attenuates and distorts ultrasound strongly and unpredictably. To address this lingering barrier, we have developed a “Relative Through-Transmit” (RTT) approach that directly measures and compensates for the attenuation and distortion of a given skull and scalp. We have implemented RTT in hardware and demonstrated that it accurately restores the operator’s intended intensities inside ex-vivo human skulls. Moreover, this functionality enabled effective and intensity-dependent transcranial modulation of nerves and effective release of defined doses of propofol inside the skull. RTT was essential for these new applications of transcranial ultrasound; when not applied, there were no significant differences from sham conditions. Moreover, RTT was safely applied in humans and accounted for all intervening obstacles including hair and ultrasound coupling. This method and hardware unlock the potential of ultrasound-based approaches to provide effective, safe, and reproducible precision therapies of the brain.
Abstract via europepmc.
Exposures
Exposure 1: Peripheral nerve (thumb) stimulation through ex-vivo skull with RTT correction
Target: digital nerve — “nerves in the thumb”
Device: custom-built
| Waveform | continuous | |
|---|---|---|
| Fundamental frequency (kHz) | 650 | ✓✓✓ |
| Pulse duration (ms) | not applicable | |
| Pulse repetition frequency (Hz) | not applicable | |
| Duty cycle (%) | 100 | ✓? |
| Sonication duration (s) | 0.3 | ✓✓✓ |
| Free-field pressure (kPa) | 1,300, 1,550, 1,800swept | ✓✓✓ |
|---|---|---|
| Free-field Isppa (W/cm²) | not reported | |
| Free-field Ispta (W/cm²) | not reported | |
| In-situ estimate | not reported | |
| In-situ pressure (kPa) | not reported | |
| In-situ Isppa (W/cm²) | not reported | |
| In-situ Ispta (W/cm²) | not reported |
Each subject experienced eight distinct stimuli, presented randomly. The ultrasound stimuli were 300 ms in duration. There were three different correction methods-hydrophone correction, no correction, and RTT correction (see above)-and a sham condition. In the sham condition, which was specifically presented for the ideal, hydrophone correction, the ultrasound was programmatically steered 10 mm below the target. For hydrophone- and RTT-corrected stimuli, we varied the intended peak pressure levels across 1.3, 1.55, and 1.8 MPa. The no correction and sham stimuli were tested only at the highest intended pressure level. We performed 10 repetitions of each stimuli producing a total of 80 trials per subject. The stimuli were delivered every 8-12 seconds and randomized so that subjects could not anticipate their onset or type.
Exposure 2: RTT through-transmit calibration scan through the human head
Target: whole brain or unfocused — “through both sides of the human head (skull attenuation measurement)”
Device: custom-built · Verasonics (driving system: Vantage256) ✓
| Waveform | pulsed | |
|---|---|---|
| Fundamental frequency (kHz) | 650 | ✓✓✓ |
| Pulse duration (ms) | 0.0154 | ✓✓✓⚑ |
| Pulse repetition frequency (Hz) | not reported | |
| Duty cycle (%) | not reported | |
| Sonication duration (s) | not reported |
| Free-field pressure (kPa) | 80 | ✓✓✓ |
|---|---|---|
| Free-field Isppa (W/cm²) | not reported | |
| Free-field Ispta (W/cm²) | not reported | |
| In-situ estimate | not reported | |
| In-situ pressure (kPa) | not reported | |
| In-situ Isppa (W/cm²) | not reported | |
| In-situ Ispta (W/cm²) | not reported | |
| Isppa, domain unspecified (W/cm²) | 1.3 | ✓✓✓ |
| Ispta, domain unspecified (W/cm²) | 0.0054 | ✓✓✓ |
The RTT scan consists of brief (< 100us) low-intensity (average peak pressure of 80 kPa in free field; Suppl. Fig. 4) pulses of ultrasound. The RTT scan takes less than one second to complete. In this method, the transducers emitted a 10-cycle, 650 kHz pulse from each of its elements while recording responses from all the other, non-transmitting elements. During the through-transmit scans, the peak pressure amplitude of each transducers was 80 kPa. The entire process of this scan takes less than 1 s to complete.
Flags from extraction
n_subjects— Paper reports two separate, non-overlapping subject groups (11 for peripheral nerve stimulation via ex-vivo skull; 5 for RTT scans through the living human head) rather than a single combined total; recorded as a list of group sizes.exposures[0]— This is a technology-validation (RTT) paper rather than a neuromodulation study; the nerve-stimulation exposure is delivered through an ex-vivo human skull to a live subject's thumb, not to the brain itself. Included because the paper explicitly frames this as ultrasonic stimulation of nerves through the skull.exposures[0].in_situ— Paper reports only relative percentages of the free-field intensity delivered into the target after RTT correction (e.g. 98.8% ± 17.8%), not an absolute in-situ pressure/intensity value, so in-situ fields are left not_reported.exposures[1]— The RTT through-transmit scan is a diagnostic-imaging-like calibration/safety measurement (not intended neuromodulation); included as a second exposure because it is explicitly applied through the living human head at a stated frequency and pressure, distinct in target and parameters from exposure 1.exposures[1].timing.pulse_duration_ms— Paper states each RTT pulse is '10-cycle, 650 kHz' and separately that scan pulses are '<100 microseconds'; converting cycles/frequency gives ~0.015 ms, but this was left not_reported rather than computed because the qualitative '<100us' description is what the paper emphasizes for safety and the two descriptions were not reconciled.readout_timing— Classified as 'online' because subjects verbally reported a percept immediately following each brief (300 ms) stimulus; paper does not use the terms 'online'/'offline'.