Magnetoelectric network spinal stimulators, 2025
Applications
Distributed magnetoelectric spinal stimulators
The August 28, 2025 published paper reports a platform with several distinct demonstrations: power-transfer measurements, an LED-addressing network, spinal stimulation and cardiac pacing. This device record covers its spinal implantable pulse generators (IPGs), not every configuration under one geometry. The acute pig application keeps the animal results separate.
This uses off-the-shelf circuitry and an ME laminate. It is not the 2022 ME-BIT ASIC/endovascular package.
The Rice magnetoelectric bioelectronics lab brief links this work to source-grounded faculty and laboratory context.
Spinal configuration
| Feature | Published spinal configuration |
|---|---|
| Packaged IPG | Approximately 1 × 1 × 1 cm per node |
| Components | Rectification, storage capacitance, programmable boost converter, microcontroller and output switch |
| ME platform films | 7.5 × 3-mm laminates;267-µm PZT between two 25-µm Metglas layers;220-kHz resonance |
| Stimulation | Voltage-controlled, 250-µs pulses, up to 14.5 V |
| Addressing | 3-bit node ID, up to eight IDs; six devices built for bench demonstration |
| Lead contacts per node | One pair on a commercial percutaneous SCS lead |
| Contact geometry | Adjacent stainless-steel cylinders, 1.33-mm diameter, 3-mm length |
| Temporary enclosure | 3D-printed box and epoxy, not proven chronic hermetic packaging |
Only two then four devices were placed in the two animal experiments. The paper’s six-device bench trace, 12-node LED demonstration and six-film power experiment are not six or 12 implanted spinal nodes.
Power and communication boundaries
The shared transmitter broadcasts commands; only the addressed node changes settings. Communication uses on-off keying with downlink rates up to 4 kbps. This paper does not demonstrate a high-bandwidth neural-recording uplink or a distributed closed-loop decoder.
The platform’s six-film measurement increases summed system efficiency from 0.22% to 1.3%, with 2.2 mW per node at 1 cm. That is summed receiver power divided by transmitter power, not improvement in each receiver’s individual efficiency or the pig spinal implant power.
The 12-node LED example is visualization of programmed outputs. A simulated 50-film network and 7.1% efficiency are simulation, not a fabricated animal network. Close-film coupling also depends on position and orientation; approximately 80% of laser-cut films passed the stated open-circuit-voltage criterion, not a guaranteed fabrication yield.
Limits
Low power-transfer efficiency still burdens the external transmitter and requires thermal/exposure management. The spinal bench setup’s 7-W transmitter demand partly reflects off-the-shelf inrush current. A battery-free implant does not make the external transmitter battery-free.
Lead placement changed activation thresholds between animal experiments. Lead-containing PZT needs a barrier against tissue/biofluid exposure. Prior 30-day glass-encapsulated work cited by the paper is not a 30-day survival result for this temporary epoxy-packaged spinal network.
Model boundary
No full model is supplied. The approximately 1-cm enclosure and contact dimensions do not recover PCB placement, film placement, wire routing, lead spacing or an exact assembly drawing.
Primary sources
- Published2025 paper.
- Full primary manuscript archive,Figure3 and spinal Methods.
- Rice institutional report,used only as program context,not additional animal data.