Research Update — October 2026

Why this matters

Nearly all transcranial ultrasound research is on adults. For children aged 0–5, the basic geometry of the skull — how thick it is, how much it varies across the width of an ultrasound probe — had not been measured in a form usable for acoustic modelling. That gap is now partly closed.

What we measured

Using the open-access Pedi-Cranial-CT-Healthy collection from the Cancer Imaging Archive (CC BY 4.0), we measured skull thickness across a 19 mm transducer aperture in 12 healthy pediatric CT scans — ten infants under one year, two children aged 3–4.

Infant skulls averaged 3.9 mm thick (range 2.4–5.6 mm). Three-year-olds measured 5.8–6.0 mm.

Skull thickness over a 19 mm aperture. Infants: mean 3.9 mm, range 2.4 to 5.6 mm. Children aged 3 to 4: 5.8 to 6.0 mm. — Figure 1. Mean skull thickness over a 19 mm aperture, 12 healthy pediatric CT scans. Infants in blue; children aged 3–4 in purple.

How much the skull varies across the beam

This is the quantity that actually distorts an ultrasound beam, and it had never been measured for children. Infant skulls vary by 0.2–0.6 mm across the aperture — smaller than earlier models assumed.

Lateral skull-thickness variation. Infant skulls vary by 0.2 to 0.6 mm across a 19 mm aperture. — Figure 2. Lateral thickness variation across the 19 mm aperture. Infant skulls: 0.2–0.6 mm.

What the beam does through a real infant skull

We fed the measured geometry into k-Wave, the standard open-source acoustic solver used across transcranial ultrasound research, and propagated a 1 MHz beam. Through a measured infant skull the beam keeps 10.9% of its energy on target against 12.4% with no skull at all — a loss of 1.5 points. The water-only control reproduced geometric arrival times to within 0.17%.

Beam energy on a 3 mm target: 12.4 percent with no skull, 10.9 percent through a measured infant skull. Simulation only. — Figure 3. Beam energy on a 3 mm target through measured infant geometry, against no skull. Simulation only.

What this means

Pediatric skulls distort the beam only mildly. The young skull is acoustically cooperative, which means the hard problem for pediatric brain assessment is detecting a faint functional signal once the beam is there — not getting energy through bone.

What is measured, and what is still assumed

Measured from real CT: skull thickness, lateral variation, and the spatial scale of that variation.
Still assumed: the acoustic properties of pediatric bone. Published CT-to-acoustic relationships were calibrated on adult bone, and no pediatric calibration exists.

Scope

Two-dimensional, lossless simulation at 1 MHz. No patient data, no phantom data, no clinical claim.

Next step

Pediatric acoustic calibration is the remaining gap. We welcome inquiries from research groups with access to relevant data or measurement facilities. Contact.

September’s Stage B, on the system page, used an analytic three-layer plate, not these CT measurements. The two results are not the same skull. Source anatomy: Pedi-Cranial-CT-Healthy, The Cancer Imaging Archive, CC BY 4.0.