Multi-Layer Brain-Mimicking Phantom for Replicating Dura and Pia Membrane Dimpling and Rupture Properties During Neural Interface Implantation

Abstract Development of novel neural interfaces faces buckling challenges and heavily relies on trial-and-error tests via in vivo animal brain insertions for design optimizations towards the minimal-damaging version for enhanced recording and stimulation outcome. To enable low-cost and fast-turnaround neural interface development and to enable previously impossible insertions via new understanding of the cutting process, this study developed a reproducible, multi-layer brain-mimicking phantom designed to replicate the rodent pia and dura mater dimpling and rupture force performance observed during in vivo tests. The phantom was composed of a 0.5% (w/v) agarose cortex layer, a 1.01% (w/v) agarose pia mater layer, and a pre-stretched polyvinyl chloride (PVC) dura mater layer, assembled via easily duplicable benchtop protocols. Using a cantilever-beam force measurement system, rupture force and dimpling depth were quantified across microwires of varying diameters (12–100 μm), materials (tungsten, stainless steel), and tip geometries, as well as segmented silicon probe shanks. Phantom test results closely matched in vivo Sprague–Dawley rat data, validating the performance of the developed multi-layer phantom. At the same time, phantom insertion trial variability was substantially lower than in vivo tests, enabling a repeatable, low-cost, early-stage screening platform of novel electrode designs. The phantom’s modular design also allowed tuning of layer thickness and stifness of each layer for diferent species or devices, ofering a validated customizable testing platform to accelerate novel neural implant development and reduce animal use.