Whole-skull acoustic transparency from a single time-reversal solve for reciprocity-based transducer placement and aperture optimization in transcranial focused ultrasound
Abstract
Transcranial focused ultrasound is limited by the skull, whose thickness, density and curvature aberrate and attenuate the beam in a position- and target-dependent way, making the choice of where to couple a transducer for a given deep target a central planning problem. This work shows that a single full-wave time-reversal solve resolves it. A virtual point source at the target radiates outward through a CT skull model, and by acoustic reciprocity the time-reversed field gives the transmit coupling of every point on the skull surface at once (a per-patch skull transparency map) and, in the same pass, the per-element signals of any array. For a phased array, time reversal conjugates the skull aberration, so the optimal placement maximizes the delivered energy, a surface integral over the aperture; a single element instead follows a coherent score mixing delivered energy with phase coherence, and the two optimizers select different windows. Placement, aperture size and orientation are then optimized by searching this single recorded map in seconds, with no per-candidate wave solve. The map is computed on a dry-skull micro-CT model for the left dentate nucleus with a 1 MHz heterogeneous Fullwave 2 solver at 6.16 points per wavelength with bone attenuation. A 120 degree occipital array refocuses within 0.25 mm of the dentate at 20.7 Pa per 1 Pa of per-element drive, and a 64 mm spherical transducer seated in the delivered-energy-optimal suboccipital window (27 mm from the target) refocuses at 7.9x gain; the same procedure focuses the thalamus at 15.2x and the dorsal anterior cingulate at 11.4x. The aberration correction raises the on-target peak 7.5x in pressure (56x in intensity) over geometric focusing, which mis-steers the focus by 4.3 mm. The same field measures the aberration's few-millimeter coherence length, which sets the required placement and target-localization accuracy.
Related
- Metalens-transparent ultrasound transducer module for extended depth photoacoustic microscopy
- MRI-Free CT-to-Ultrasound Skull Registration in Transcranial Focused Ultrasound using CMUT based Integrated Therapeutic and Imaging Arrays
- Spatiotemporal Adaptive Eigenspectrum Remapping in Transcranial Ultrasound Localization Microscopy
- Self-adaptive transcranial ultrasound imaging using an effective skull model: experimental evaluation on a human skull at 3.15 MHz
- Earth as a transducer for ultralight bosonic dark-matter detection
BibTeX
@misc{pinton2026wholeskull,
title = {{Whole-skull acoustic transparency from a single time-reversal solve for reciprocity-based transducer placement and aperture optimization in transcranial focused ultrasound}},
author = {Gianmarco Pinton},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.27852},
eprint = {2609.27852},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.27852},
}
RIS
TY - UNPB
TI - Whole-skull acoustic transparency from a single time-reversal solve for reciprocity-based transducer placement and aperture optimization in transcranial focused ultrasound
AU - Gianmarco Pinton
PY - 2026
DA - 2026/09/01
PB - arXiv
DO - 10.48550/arxiv.2609.27852
UR - https://arxiv.org/abs/2609.27852
AB - Transcranial focused ultrasound is limited by the skull, whose thickness, density and curvature aberrate and attenuate the beam in a position- and target-dependent way, making the choice of where to couple a transducer for a given deep target a central planning problem. This work shows that a single full-wave time-reversal solve resolves it. A virtual point source at the target radiates outward through a CT skull model, and by acoustic reciprocity the time-reversed field gives the transmit coupling of every point on the skull surface at once (a per-patch skull transparency map) and, in the same pass, the per-element signals of any array. For a phased array, time reversal conjugates the skull aberration, so the optimal placement maximizes the delivered energy, a surface integral over the aperture; a single element instead follows a coherent score mixing delivered energy with phase coherence, and the two optimizers select different windows. Placement, aperture size and orientation are then optimized by searching this single recorded map in seconds, with no per-candidate wave solve. The map is computed on a dry-skull micro-CT model for the left dentate nucleus with a 1 MHz heterogeneous Fullwave 2 solver at 6.16 points per wavelength with bone attenuation. A 120 degree occipital array refocuses within 0.25 mm of the dentate at 20.7 Pa per 1 Pa of per-element drive, and a 64 mm spherical transducer seated in the delivered-energy-optimal suboccipital window (27 mm from the target) refocuses at 7.9x gain; the same procedure focuses the thalamus at 15.2x and the dorsal anterior cingulate at 11.4x. The aberration correction raises the on-target peak 7.5x in pressure (56x in intensity) over geometric focusing, which mis-steers the focus by 4.3 mm. The same field measures the aberration's few-millimeter coherence length, which sets the required placement and target-localization accuracy.
ER -