The implementation of the CLEPSYDRA system will be developed along four areas of activity, each including design, prototyping and measurement tasks (a more detailed description of each item can be found in the Research
- Reconfigurable neural recording system (mixed-signal implant SoC)
- Exploration, analysis and design of the signal acquisition and conditioning circuitry for the neural activity recorded in the electrodes.
- Exploration, analysis and design of data mining mechanisms for the on-the-fly compression of neural recordings.
- Definition and design of the circuits and protocols for the single-wire communication interface between the stacked SoCs.
- Digital processor SoC for epilepsy prognosis (digital implant SoC)
- Development of an on-chip artifact removal device which can remove the contaminant waveforms resulting neural stimulation or line interference.
- Implementation of a multi-channel real-time online spike sorter that can dynamically determine the number of neuron clusters, and store their relevant features.
- Exploration, analysis and design of a real-time multivariate seizure predictor based on the data reduced in the previous processing stages.
- Telemetry and power management (RF link implant SoC)
- Exploration, analysis and design of a power/data telemetry unit based on inductive coupling.
- Implementation of a mixed-external external unit using off-the-shelves components for remotely controlling the complete neural prostheses.
- System integration and assembly
- Optimum design of platforms and testing protocols for the characterization of the fabricated SoCs, either individually, or connected to another via stacked-die wire bonding.
- Building the set-up for in vitro measurements.
- Building the set-up for in vivo measurements.