Body Area Networks: Smart IoT and Big Data for Intelligent Health Management. 14th EAI International Conference, BODYNETS 2019, Florence, Italy, October 2-3, 2019, Proceedings

Research Article

Arrhythmia Detection with Antidictionary Coding and Its Application on Mobile Platforms

  • @INPROCEEDINGS{10.1007/978-3-030-34833-5_5,
        author={Gilson Frias and Hiroyoshi Morita and Takahiro Ota},
        title={Arrhythmia Detection with Antidictionary Coding and Its Application on Mobile Platforms},
        proceedings={Body Area Networks:  Smart IoT and Big Data for Intelligent Health Management. 14th EAI International Conference, BODYNETS 2019, Florence, Italy, October 2-3, 2019, Proceedings},
        proceedings_a={BODYNETS},
        year={2019},
        month={11},
        keywords={ECG Arrhythmia Antidictionary Werable Mobile},
        doi={10.1007/978-3-030-34833-5_5}
    }
    
  • Gilson Frias
    Hiroyoshi Morita
    Takahiro Ota
    Year: 2019
    Arrhythmia Detection with Antidictionary Coding and Its Application on Mobile Platforms
    BODYNETS
    Springer
    DOI: 10.1007/978-3-030-34833-5_5
Gilson Frias1,*, Hiroyoshi Morita1,*, Takahiro Ota2,*
  • 1: The University of Electro-Communications
  • 2: Nagano Prefectural Institute of Technology
*Contact email: gilson.frias@mail.uec.jp, morita@uec.ac.jp, ota@cse.pit-nagano.ac.jp

Abstract

In response to the demand of memory efficient algorithms for electrocardiogram (ECG) signal processing and anomaly detection on wearable and mobile devices, an implementation of the antidictionary coding algorithm for memory constrained devices is presented. Pre-trained probabilistic models built from quantized ECG sequences were constructed in an offline fashion and their performance was evaluated on a set of test signals. The low complexity requirements of the models is confirmed with a port of a pre-trained model of the algorithm into a mobile device without incurring on excessive use of computational resources.