Highly-Digital 0.0018-mm2/channel Multiplexed Neural Frontend with Time-Based Incremental ADC for In-Brain-Stroke-Cavity LFP Monitoring

Abnormal brain cavities are of primary interest in the treatment of stroke symptoms and traumatic brain injuries because of the highly plastic behavior of the neurons along the cavity wall. Their complex and hollow geometry, however, renders the use of rigid needle probes or traditional μECoG arrays ineffective. The deeply-seated implant also dictates strict area and heat dissipation constraints that most published designs do not meet. This paper presents the design of an ultra-compact neural frontend that is part of the fully implantable SCATMAN system, tailored for in-cavity Local Field Potential (LFP) monitoring. The prototype has been fabricated in a 180-nm bulk CMOS technology. To fit the cavity size constraints, this design combines time multiplexing and time-domain conversion in a novel 13-bit incremental ADC, requiring only 0.0018 mm
²
/channel of readout area, the smallest reported for stimulation-voltage-compatible technologies, while consuming 4.51 μW/channel. The use of a time-based ∆Σ modulator renders the implementation highly digital and ensures a very small silicon footprint. The design also yields a best-in-class peak Total Harmonic Distortion (THD) of -72.6 dB @2.5-mVpp input, thanks to its implicit DAC mismatch-error shaping property.