A fully-integrated analog machine learning classifier for breast cancer classification

Sanjeev T. Chandrasekaran; Ruobing Hua; Imon Banerjee; Arindam Sanyal

doi:10.3390/electronics9030515

A fully-integrated analog machine learning classifier for breast cancer classification

Sanjeev T. Chandrasekaran, Ruobing Hua, Imon Banerjee, Arindam Sanyal

Diagnostic Radiology

Research output: Contribution to journal › Article › peer-review

Abstract

We propose a fully integrated common-source amplifier based analog artificial neural network (ANN). The performance of the proposed ANN with a custom non-linear activation function is demonstrated on the breast cancer classification task. A hardware-software co-design methodology is adopted to ensure good matching between the software AI model and hardware prototype. A 65 nm prototype of the proposed ANN is fabricated and characterized. The prototype ANN achieves 97% classification accuracy when operating from a 1.1 V supply with an energy consumption of 160 fJ/classification. The prototype consumes 50 µW power and occupies 0.003 mm² die area.

Original language	English (US)
Article number	515
Journal	Electronics (Switzerland)
Volume	9
Issue number	3
DOIs	https://doi.org/10.3390/electronics9030515
State	Published - Mar 2020

Keywords

Analog-AI
Breast cancer detection
CS amplifier
Classification
Intelligence-at-the-edge
Machine learning
Neural network

ASJC Scopus subject areas

Control and Systems Engineering
Signal Processing
Hardware and Architecture
Computer Networks and Communications
Electrical and Electronic Engineering

Access to Document

10.3390/electronics9030515

Cite this

@article{183337ed7fd64b3d8b31cd45979bc0df,

title = "A fully-integrated analog machine learning classifier for breast cancer classification",

abstract = "We propose a fully integrated common-source amplifier based analog artificial neural network (ANN). The performance of the proposed ANN with a custom non-linear activation function is demonstrated on the breast cancer classification task. A hardware-software co-design methodology is adopted to ensure good matching between the software AI model and hardware prototype. A 65 nm prototype of the proposed ANN is fabricated and characterized. The prototype ANN achieves 97% classification accuracy when operating from a 1.1 V supply with an energy consumption of 160 fJ/classification. The prototype consumes 50 µW power and occupies 0.003 mm2 die area.",

keywords = "Analog-AI, Breast cancer detection, CS amplifier, Classification, Intelligence-at-the-edge, Machine learning, Neural network",

author = "Chandrasekaran, {Sanjeev T.} and Ruobing Hua and Imon Banerjee and Arindam Sanyal",

note = "Publisher Copyright: {\textcopyright} 2020 by the authors. Licensee MDPI, Basel, Switzerland.",

year = "2020",

month = mar,

doi = "10.3390/electronics9030515",

language = "English (US)",

volume = "9",

journal = "Electronics (Switzerland)",

issn = "2079-9292",

publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",

number = "3",

}

TY - JOUR

T1 - A fully-integrated analog machine learning classifier for breast cancer classification

AU - Chandrasekaran, Sanjeev T.

AU - Hua, Ruobing

AU - Banerjee, Imon

AU - Sanyal, Arindam

PY - 2020/3

Y1 - 2020/3

N2 - We propose a fully integrated common-source amplifier based analog artificial neural network (ANN). The performance of the proposed ANN with a custom non-linear activation function is demonstrated on the breast cancer classification task. A hardware-software co-design methodology is adopted to ensure good matching between the software AI model and hardware prototype. A 65 nm prototype of the proposed ANN is fabricated and characterized. The prototype ANN achieves 97% classification accuracy when operating from a 1.1 V supply with an energy consumption of 160 fJ/classification. The prototype consumes 50 µW power and occupies 0.003 mm2 die area.

AB - We propose a fully integrated common-source amplifier based analog artificial neural network (ANN). The performance of the proposed ANN with a custom non-linear activation function is demonstrated on the breast cancer classification task. A hardware-software co-design methodology is adopted to ensure good matching between the software AI model and hardware prototype. A 65 nm prototype of the proposed ANN is fabricated and characterized. The prototype ANN achieves 97% classification accuracy when operating from a 1.1 V supply with an energy consumption of 160 fJ/classification. The prototype consumes 50 µW power and occupies 0.003 mm2 die area.

KW - Analog-AI

KW - Breast cancer detection

KW - CS amplifier

KW - Classification

KW - Intelligence-at-the-edge

KW - Machine learning

KW - Neural network

UR - http://www.scopus.com/inward/record.url?scp=85083264602&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85083264602&partnerID=8YFLogxK

U2 - 10.3390/electronics9030515

DO - 10.3390/electronics9030515

M3 - Article

AN - SCOPUS:85083264602

SN - 2079-9292

VL - 9

JO - Electronics (Switzerland)

JF - Electronics (Switzerland)

IS - 3

M1 - 515

ER -

A fully-integrated analog machine learning classifier for breast cancer classification

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this