A seizure-detection IC employing machine learning to overcome data-conversion and analog-processing non-idealities

Jintao Zhang, Liechao Huang, Zhuo Wang, Naveen Verma

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Scopus citations

Abstract

This paper presents a seizure-detection system wherein the accuracy required of the analog frontend is substantially relaxed. Typically, readout of electroencephalogram (EEG) signals would dominate the energy of such a system, due to the precision (noise, linearity) requirements. The presented system performs data conversion and analog multiplication for EEG feature extraction via simple circuits to demonstrate that feature errors can be overcome by appropriate retraining of a classification model, using a machine-learning algorithm. This precludes the need to design a high-precision frontend. The prototype, in 32nm CMOS, results in features whose RMS error normalized to their ideal values is 1.16 (i.e. errors are larger than ideal values). An ideal implementation of the seizure detector exhibits sensitivity, latency, false alarms of 5/5, 2.0 sec., 8, respectively. The feature errors degrade this to 5/5, 3.6 sec., 443, causing high false alarms; but retraining of the classification model restores this to 5/5, 3.4 sec., 4.

Original languageEnglish (US)
Title of host publication2015 IEEE Custom Integrated Circuits Conference, CICC 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781479986828
DOIs
StatePublished - Nov 25 2015
EventIEEE Custom Integrated Circuits Conference, CICC 2015 - San Jose, United States
Duration: Sep 28 2015Sep 30 2015

Publication series

NameProceedings of the Custom Integrated Circuits Conference
Volume2015-November
ISSN (Print)0886-5930

Other

OtherIEEE Custom Integrated Circuits Conference, CICC 2015
CountryUnited States
CitySan Jose
Period9/28/159/30/15

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering

Keywords

  • Machine learning
  • analog processing circuits
  • electroencephalography
  • epilepsy
  • error compensation
  • system-on-chip

Fingerprint Dive into the research topics of 'A seizure-detection IC employing machine learning to overcome data-conversion and analog-processing non-idealities'. Together they form a unique fingerprint.

  • Cite this

    Zhang, J., Huang, L., Wang, Z., & Verma, N. (2015). A seizure-detection IC employing machine learning to overcome data-conversion and analog-processing non-idealities. In 2015 IEEE Custom Integrated Circuits Conference, CICC 2015 [7338456] (Proceedings of the Custom Integrated Circuits Conference; Vol. 2015-November). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CICC.2015.7338456