Drawbacks of Current State-of-the-Art Deep Brain Stimulation Architectures
Deep brain stimulation (DBS) of the sub-thalamic nucleus is an effective treatment for numerous neurological disorders such as tremor, dystonia, and Parkinson’s disease. However, despite much research, algorithms and systems for closed loop operation remain elusive. The spike sorting and principal component analysis (PCA) approach holds promise but requires significant computation, making radio-frequency (RF) or long-term battery powered operation difficult.
University of Michigan’s Deep Brain Stimulation SoC Architecture
Researchers at the University of Michigan have developed a single-chip log-based wirelessly-powered closed-loop DBS system with RF telemetry that overcomes the issues prevalent in current state-of-the-art designs.
The system-on-a-chip (SoC) architecture makes use of the energy of Low Frequency Field Potentials (LFPs), a more effective feedback indicator for DBS than the traditional spike sorting and PCA approach. The system optimizes stimulation parameters by analyzing measured LFP energy, and incorporates extensive two-way telemetry, and filtering of spike and LFP signals. Since neural signals are expressed and processed in the log-domain, substantial energy savings can be realized. Further, aside from certain high-voltage components, the entire system is powered using energy recovered from a carrier signal by an efficient RF to DC converter.
Applications and Advantages
- Deep brain stimulation (DBS) for treating neurological disorders like
- Parkinson’s disease
- Standalone and unified system-on-a-chip (SoC) architecture
- Closed loop operation
- Low Frequency Field Potential (LFP) energy based design
- Low power requirements
- Wirelessly powered
- Powered by energy harvesting
- Log domain processing
- Support for electrical and optical stimulation (for optogenetic research)
- Radio telemetry support
- Large effective dynamic range