Back to search

Cornell University

Scalable, Low-Power Integrated Circuits and Systems for Future Intelligent Neural Interfaces

Abstract

dc:description.abstract

Neurological disorders are the second leading cause of global deaths and the leading cause of disability worldwide. Tackling this challenge has spurred extensive research efforts to develop intelligent neural implants capable of high-density neural sensing, artificial intelligence (AI)-driven brain-state classification, and therapeutic feedback via neurostimulation. Similarly, prosthetic brain-machine interface (BMI) technology has recently experienced enormous progress, leveraging complementary metal-oxide-semiconductor (CMOS) and AI technology to provide unprecedented opportunities for individuals with paralysis or limb loss. However, integrating these intricate functions into a single system faces significant challenges, primarily due to the stringent power and area budgets of battery-operated neural implants. Therefore, this dissertation aims to overcome these bottlenecks through innovative CMOS integrated circuit (IC) and algorithm co-design, paving the way for next-generation intelligent neural interfaces. Chapter 2 presents a 256-channel intelligent neural system-on-chip (SoC) that notably enhances the sensing channel count through a highly scalable sensor front-end architecture. An energy-efficient, versatile machine learning processor is integrated to facilitate brain-state classification across various neurological conditions. In Chapter 3, I describe a pioneering phase-locked deep brain stimulation (DBS) SoC that can monitor and regulate abnormal brain connectivity in network-based brain disorders, such as depression. A novel phase-locked DBS scheme is introduced and demonstrated in animal experiments for the first time in literature. Finally, Chapter 4 discusses a 192-channel compact, low-power neural recording IC for prosthetic BMI applications. The recording IC is further integrated with an AI-enhanced neural decoder in a miniaturized BMI chipset enabling 31-class handwriting intention decoding from human brain activity.

Degree

thesis:*
Name thesis:degree_name
Ph. D., Electrical and Computer Engineering
Level thesis:degree_level
Doctor of Philosophy
Discipline thesis:degree_discipline
Electrical and Computer Engineering
Grantor
Cornell University
Year dc:date.issued
2024

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Shin, Uisub
Chair dc:contributor.chair
  • Shoaran, Mahsa
Committee members dc:contributor.committeemember
  • Zhang, Zhiru
  • Goldberg, Jesse

Subjects

dc:subject × 6

Rights

Language dc:language.iso
en

Identifiers

dc:identifier.*
Dc Identifier Bibid
16575423
Dc Identifier Other
ProQuest Submission ID: 14273
ProQuest Publication ID: 31242724

Chain of custody

source
Harvested from
Cornell University
Base URL
ecommons.cornell.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Shin, Uisub. Scalable, Low-Power Integrated Circuits and Systems for Future Intelligent Neural Interfaces. Doctor of Philosophy thesis, Cornell University, 2024. https://hdl.handle.net/1813/115999