Neural Noise- The efficacy of neuromodulation techniques like TMS is highly dependent on the geometry and conductivity of the stimulated brain target.1
- Anatomically and conductively accurate rat head phantoms enable measurement of TMS-induced electric fields for stimulation targeting.1 1
Weekly enrichment (2026-07-20)
- Lohr et al. (2026, J. Neural Eng. 23:016028) built anatomically accurate rat head phantoms via 3D printing and injection molding, using public MRI/CT data and the SIGMA rat brain atlas segmented in SPM12 and 3D Slicer to generate skin, skull, CSF, gray-matter, and white-matter compartments.2
- The conductive brain surrogate was a polydimethylsiloxane (PDMS) composite loaded with multiwalled carbon nanotubes (20–30 µm long, 80 nm wide); a 10.35 wt% CNT loading reproduced the target brain conductivity of ~0.5 S m⁻¹ (reference gray matter 0.41 S m⁻¹, white matter 0.37 S m⁻¹).2
- Induced E-fields were read out with embedded mutually orthogonal triaxial dipole probes (TDP) of 37-gauge copper wire, with ~3 mm dipole spacing and twisted leads to cancel magnetic pickup, positioned at brain-surface depths of 3, 6, and 7 mm from the coil.2
- Four hand-wound figure-of-eight coils (inductance 10.2–15.9 µH) were tested on a MagVenture MagPro R30 (max output 5000 A, ~2.5 kHz waveform) at 5%, 10%, and 15% output (≈333, 666, 1000 A), each with V-tip or flat-tip cores of iron-cobalt-vanadium (permendur) or AISI 1010 carbon steel.2
- FEM simulations in Sim4Life (v6.2.1.4972) with IT’IS LF v4.0 tissue properties matched measurements with an average error of 5.1%; peak measured-vs-simulated surface E-fields were 115.3–110 V m⁻¹ (permendur V-tip), 91.9–85 V m⁻¹ (permendur flat-tip), 94.7–100 V m⁻¹ (AISI 1010 V-tip), and 85.9–84 V m⁻¹ (AISI 1010 flat-tip).2
- The permendur V-tip core gave the highest and most focal stimulation, and homogenizing gray/white matter into one brain compartment had negligible effect on the field profile (Cohen’s d = 0.0023 between multilayer and homogeneous simulations).2
- The authors position these phantoms as a way to validate coils and protocols before in-vivo work, reducing dependence on IACUC-approved animal experiments, and propose machine-learning models trained on FEM data to reconstruct whole-brain E-field profiles from sparse probe sites; the underlying data are shared openly at OSF.2
- The approach fits a broader preclinical dosimetry literature validating Sim4Life rodent TMS simulations against physical measurements, including prior work on soft-magnetic core shapes and high-performance magnetic-core coils for targeted rodent brain stimulation.3