Related Experiment Video
Updated: May 14, 2026
11:54
Microfluidic Platform with Multiplexed Electronic Detection for Spatial Tracking of Particles
Published on: March 13, 2017
Low power and high accuracy spike sorting microprocessor with on-line interpolation and re-alignment in 90 nm CMOS
Tung-Chien Chen1, Tsung-Chuan Ma, Yun-Yu Chen
1DSP/IC Design Lab, Graduate Institute of Electronics Engineering, National Taiwan University, Taiwan. djchen@video.ee.ntu.edu.tw
Summary
This study introduces on-chip spike sorting with integrated interpolation hardware. This innovation improves accuracy and power efficiency, allowing lower sampling rates without performance compromise in neural recordings.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Integrated Circuit Design
Background:
- Accurate spike sorting is crucial for neuroscience and neuroprosthetics.
- Higher sampling rates (SR) improve spike sorting accuracy but increase power consumption.
- Miniaturized, wireless neural recording ICs for free-moving subjects often require lower SR due to power constraints.
Purpose of the Study:
- To develop an on-chip spike sorting processor with integrated interpolation hardware.
- To enhance the trade-off between power consumption and accuracy in neural recording systems.
- To enable effective spike sorting at lower sampling rates.
Main Methods:
- Implementation of an on-chip spike sorting processor.
- Integration of dedicated hardware for waveform interpolation.
- Fabrication using a 90nm process technology.
Main Results:
- The system with integrated interpolation achieved superior accuracy and power efficiency compared to a system without interpolation at a higher SR.
- A system operating at 12.5 k samples per second (sps) with interpolation outperformed a 25 ksps system without interpolation.
- Demonstrated improved performance in terms of both accuracy and power consumption.
Conclusions:
- Integrated interpolation hardware can significantly enhance on-chip spike sorting performance.
- This approach allows for effective neural data processing at reduced sampling rates and power consumption.
- The developed processor offers a viable solution for long-duration, low-power neural recording applications.
