Miniaturized Energy-Efficient Integrated Neural Interfaces

Chul KimDepartment of Bio and Brain Engineering

KAIST

Microelectrocorticography (µECoG)

Emerging µECoG technologies- High spatiotemporal resolution- Fully implantable

- Minimally invasive- Long-term monitoring

Slide 1

Emerging μECoG TechnologiesFlexibleArray Modular

ENIAC:Encapsulated Neural Interfacing and Acquisition Chip

No external components or wires

CICC2017, JSSC2018b, JSSC2018c, JSSC2017, VLSI2015a, VLSI2016, ProcIEEE2017, BioCAS2017Slide 12

Slide 3

Neural Recording ADC – State-of-the-Art

Large GainMax

range

Slide 4

Neural Recording ADC – Improvements

ADC-Direct Front-End

Maxrange

JSSC2018b

Neural Recording ADC – Micrograph

Slide 5ISSCC2018

Slide 6ISSCC2018, JSSC2018b

Neural Recording ADC – in vivo testsLFPs in marmoset non-human primate

Slide 7ISSCC2018, JSSC2018b

Neural Recording ADC – in vivo testsLFPs in marmoset non-human primate

Collaboration

Scalp

Skull

Dura materArachnoid

Pia mater

Outer skull table

Inner skull table

Cortex

CSF

Diploe

Flexible Array “Modular” µECoG Implants

Slide 8

1 mm

1 mm

Joint proposal opportunities for µECoG systems working on freely moving subjects for next-generation brain imaging with:

1) < 1mV spatial resolution;2) > 1000 recording channels;3) > 100 stimulation channels; 4) > at least 24 hours operation time, and;5) coverage of entire brain.

Joint proposal opportunities for unobtrusive wireless energy-efficient brain-machine interface systems.