2.5D heterogeneously integrated microsystem for high-density neural sensing applications.

Area of Science:

• Neuroscience
• Electrical Engineering
• Materials Science

Background:

• Miniaturized neural sensing microsystems are essential for investigating brain function.
• High-density neural signal acquisition requires advanced integration techniques.

Purpose of the Study:

• To present a 2.5D heterogeneously integrated bio-sensing microsystem with μ-probes and embedded through-silicon-via (TSVs).
• To enable high-density neural sensing applications with reduced power consumption.

Main Methods:

• Fabrication of a 24 × 24 μ-probe array with embedded TSVs on a 5×5 mm² chip.
• Heterogeneous integration of μ-probes, 4 dies (for signal acquisition, feature extraction, and classification), and a silicon interposer.
• Utilizing low-power analog front-end circuits, efficient analog-to-digital converters, discrete wavelet transforms, filters, and a microcontroller (MCU).
• Implementation of an on-interposer bus (μ-SPI) for data transfer.

Main Results:

• Successful fabrication and integration of the 2.5D bio-sensing microsystem.
• Demonstration of 16-channel neural signal capture with high-density μ-probes.
• Achieved a low overall power consumption of 676.3 μW for 16-channel neural sensing.
• Successful in-vivo testing validating the microsystem's performance.

Conclusions:

• The proposed 2.5D heterogeneously integrated bio-sensing microsystem is effective for high-density neural sensing.
• The microsystem offers a low-power solution for brain function investigation.
• This technology holds promise for advancing neural interface applications.