3D Printing of Flexible, Scaled Neuron Models

Abstract

Microscopy-based 3D neuronal reconstructions are freely available online, including in the NeuroMorpho.Org file repository. Each neuron’s dendritic structure is intricate and diverse, making it challenging to produce accurate physical 3D models for instruction or visualization. This work examines several methods for producing 3D models of neuronal reconstructions and compares their cost and accessibility. In response to high cost of direct 3D printing methods, we develop a new casting method which uses 3D-printed, single-use dissolvable molds and achieves lower cost for producing 3D neuron models. The casting method uses a consumer-grade desktop fused filament fabrication 3d printer, water-soluble polyvinyl alcohol filament, and a two-part casting material such as polyurethane resin or silicone rubber. Physical models of a diverse set of neuron morphologies including purkinje, pyramidal, medium spiny, and retinal ganglion cells were produced using the casting method with good fidelity to the neuronal reconstruction file and sufficient detail and strength for hands-on use in neuroscience education and research. The average cost of producing the four neuron models using the proposed casting method was reduced by 58% relative to the cost of using the least expensive 3D printing method by a service provider. Production time for one neuronal model using the proposed method was found to be in the range of 1-3 days while service-provided neurons required a minimum of a week from order placement to delivery.