Designing a high-efficiency charge balancer for implantable electrical stimulator ICs
This paper presents a mixed-signal implantable integrated circuit (IC) designed for safe and energy-efficient charge-balanced neuromuscular electrical stimulation (NMES). NMES has emerged as an effective treatment for a wide range of symptoms and disorders, including restoring motor function in partially-denervated muscles, alleviating pain, and even recovering sensory functions such as hearing and vision. While numerous implantable NMES stimulator architectures have been reported in the literature, many still suffer from limitations in silicon area, power efficiency, and safety issues related to charge accumulation. The proposed design directly targets these challenges through a compact and energy-efficient mixed-signal IC implementation. The proposed design integrates a programmable current driver and a charge-balancing feedback loop to maintain long-term stimulation safety. The stimulation current is programmable from 100 μ A to 800 μ A, providing a suitable range for neuronal stimulation, such as that of epidural spinal cord nerves. The IC achieves a compact core area of 0.1375 mm 2 , while maintaining a high-power efficiency of 66%, representing a notable improvement over previously reported designs in the literature. The circuit demonstrates robust operation across process, voltage, and temperature (PVT) variations, maintaining both high efficiency and accurate current delivery. Post-stimulation charge residuals remain within a ±35 mV safety window, ensuring safe charge-balanced stimulation and enhanced biocompatibility. Maintaining this low residual voltage is essential to prevent electrode corrosion and irreversible Faradaic reactions, which can generate toxic by-products and lead to potential of hydrogen (pH) change at the tissue–electrode interface. These results confirm the IC’s potential for enabling low-power, safe, long-term, and scalable neurostimulation systems, with the flexibility to serve as a reusable intellectual property (IP) block across a wide range of implantable clinical applications.
Institutions
- University of Mississippi (US)
- Assiut University (EG)
Publication Details
- Journal
- Computers & Electrical Engineering
- Published
- 2026-05-19
- DOI
- https://doi.org/10.1016/j.compeleceng.2026.111219
- Primary Topic
- Neuroscience and Neural Engineering
- Type
- article
- Field-Weighted Citation Impact
- 0.00