Towards Virtual Prototyping of Electrically Active Implants Using SystemC-AMS

Deep Brain Stimulation is an indispensable therapy option for severe neurodegenerative diseases like parkinson. However, the lack of understanding of the functional principle hinders an optimal clinical application. Therefore, basic research in animal models is crucial. For the development of neurostimulators in rodents, physical prototyping alone is not feasible anymore due to the growing complexity and necessary optimization of the overall implants. However, there is no scientific approach for virtual prototyping of neurostimulators. This work proposes a power model for a developed neurostimulator, based on a time continuous battery model and power-state-machines. The model was implemented in SystemC and SystemC-AMS and validated against measurements. The model helped us understand and find power optimal working points of the implant and thus achieve a runtime gain of 44% with a given battery. Consequently, this work shows that we can use virtual prototyping in SystemC-AMS to aid design decisions for future implant versions.