Serverless data parallelization for training and retraining of deep learning architecture in patient-specific arrhythmia detection

Michael Marefat, Amit Juneja

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

Abstract

Stacked Denoising Autoencoders (SDA) are deep networks which have superior generative properties and therefore can be trained and retrained to learn the structure of a patient's heart beat signal with minimal training data. This approach is particularly useful in continuous remote devices because they gather large amounts of data for longer periods of time. Serverless applications are the desired way of building applications due to its cost effectiveness after advancements in commercially available serverless host providers like Amazon AWS. This work proposes a serverless architecture for the training and retraining of SDA, for classification of arrhythmias in a patient-specific manner. This work also proposes a technique for data parallelization in the serverless architecture to achieve a speedup of up-To 13x in training time. This work uses MIT-BIH Arrhythmia database. Retraining with this architecture shows high classification accuracies for Ventricular Ectopic Beats (VEB) (97.41%) and Supraventricular Ectopic Beats (SVEB) (98.77%).

Original languageEnglish (US)
Title of host publication2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781728108483
DOIs
StatePublished - May 2019
Event2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Chicago, United States
Duration: May 19 2019May 22 2019

Publication series

Name2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings

Conference

Conference2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019
CountryUnited States
CityChicago
Period5/19/195/22/19

Fingerprint

Cardiac Arrhythmias
Learning
Ventricular Premature Complexes
Cost effectiveness
Cost-Benefit Analysis
Databases
Equipment and Supplies
Deep learning
Retraining
Amazon
Cost-effectiveness
Long period
Data base

Keywords

  • Arrhythmia Classification
  • Deep Learning
  • Patient Specific
  • Remote Continuous Health Devices
  • Retraining
  • Serverless
  • Stacked Denoising Autoencoders

ASJC Scopus subject areas

  • Artificial Intelligence
  • Signal Processing
  • Information Systems and Management
  • Biomedical Engineering
  • Health Informatics
  • Radiology Nuclear Medicine and imaging

Cite this

Marefat, M., & Juneja, A. (2019). Serverless data parallelization for training and retraining of deep learning architecture in patient-specific arrhythmia detection. In 2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings [8834566] (2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/BHI.2019.8834566

Serverless data parallelization for training and retraining of deep learning architecture in patient-specific arrhythmia detection. / Marefat, Michael; Juneja, Amit.

2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. 8834566 (2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings).

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

Marefat, M & Juneja, A 2019, Serverless data parallelization for training and retraining of deep learning architecture in patient-specific arrhythmia detection. in 2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings., 8834566, 2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings, Institute of Electrical and Electronics Engineers Inc., 2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019, Chicago, United States, 5/19/19. https://doi.org/10.1109/BHI.2019.8834566
Marefat M, Juneja A. Serverless data parallelization for training and retraining of deep learning architecture in patient-specific arrhythmia detection. In 2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2019. 8834566. (2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings). https://doi.org/10.1109/BHI.2019.8834566
Marefat, Michael ; Juneja, Amit. / Serverless data parallelization for training and retraining of deep learning architecture in patient-specific arrhythmia detection. 2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. (2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings).
@inproceedings{1809dc53812f40cf9858e9ae073877c6,
title = "Serverless data parallelization for training and retraining of deep learning architecture in patient-specific arrhythmia detection",
abstract = "Stacked Denoising Autoencoders (SDA) are deep networks which have superior generative properties and therefore can be trained and retrained to learn the structure of a patient's heart beat signal with minimal training data. This approach is particularly useful in continuous remote devices because they gather large amounts of data for longer periods of time. Serverless applications are the desired way of building applications due to its cost effectiveness after advancements in commercially available serverless host providers like Amazon AWS. This work proposes a serverless architecture for the training and retraining of SDA, for classification of arrhythmias in a patient-specific manner. This work also proposes a technique for data parallelization in the serverless architecture to achieve a speedup of up-To 13x in training time. This work uses MIT-BIH Arrhythmia database. Retraining with this architecture shows high classification accuracies for Ventricular Ectopic Beats (VEB) (97.41{\%}) and Supraventricular Ectopic Beats (SVEB) (98.77{\%}).",
keywords = "Arrhythmia Classification, Deep Learning, Patient Specific, Remote Continuous Health Devices, Retraining, Serverless, Stacked Denoising Autoencoders",
author = "Michael Marefat and Amit Juneja",
year = "2019",
month = "5",
doi = "10.1109/BHI.2019.8834566",
language = "English (US)",
series = "2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
booktitle = "2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings",

}

TY - GEN

T1 - Serverless data parallelization for training and retraining of deep learning architecture in patient-specific arrhythmia detection

AU - Marefat, Michael

AU - Juneja, Amit

PY - 2019/5

Y1 - 2019/5

N2 - Stacked Denoising Autoencoders (SDA) are deep networks which have superior generative properties and therefore can be trained and retrained to learn the structure of a patient's heart beat signal with minimal training data. This approach is particularly useful in continuous remote devices because they gather large amounts of data for longer periods of time. Serverless applications are the desired way of building applications due to its cost effectiveness after advancements in commercially available serverless host providers like Amazon AWS. This work proposes a serverless architecture for the training and retraining of SDA, for classification of arrhythmias in a patient-specific manner. This work also proposes a technique for data parallelization in the serverless architecture to achieve a speedup of up-To 13x in training time. This work uses MIT-BIH Arrhythmia database. Retraining with this architecture shows high classification accuracies for Ventricular Ectopic Beats (VEB) (97.41%) and Supraventricular Ectopic Beats (SVEB) (98.77%).

AB - Stacked Denoising Autoencoders (SDA) are deep networks which have superior generative properties and therefore can be trained and retrained to learn the structure of a patient's heart beat signal with minimal training data. This approach is particularly useful in continuous remote devices because they gather large amounts of data for longer periods of time. Serverless applications are the desired way of building applications due to its cost effectiveness after advancements in commercially available serverless host providers like Amazon AWS. This work proposes a serverless architecture for the training and retraining of SDA, for classification of arrhythmias in a patient-specific manner. This work also proposes a technique for data parallelization in the serverless architecture to achieve a speedup of up-To 13x in training time. This work uses MIT-BIH Arrhythmia database. Retraining with this architecture shows high classification accuracies for Ventricular Ectopic Beats (VEB) (97.41%) and Supraventricular Ectopic Beats (SVEB) (98.77%).

KW - Arrhythmia Classification

KW - Deep Learning

KW - Patient Specific

KW - Remote Continuous Health Devices

KW - Retraining

KW - Serverless

KW - Stacked Denoising Autoencoders

UR - http://www.scopus.com/inward/record.url?scp=85073011419&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85073011419&partnerID=8YFLogxK

U2 - 10.1109/BHI.2019.8834566

DO - 10.1109/BHI.2019.8834566

M3 - Conference contribution

AN - SCOPUS:85073011419

T3 - 2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings

BT - 2019 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2019 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

ER -