An online swallow detection algorithm based on the quadratic variation of dual-axis accelerometry

Authors:
Sotirios Damouras;Ervin Sejdić;Catriona M. Steele;Tom Chau
Affiliations:
Bloorview Research Institute, Bloorview Kids Rehab, Toronto, ON, Canada;Bloorview Research Institute, Bloorview Kids Rehab and the Institute of Biomaterials and Biomedical Engineering, University of Toronto, Toronto, ON, Canada;Toronto Rehabilitation Institute and the Department of Speech-Language Pathology, University of Toronto, Toronto, ON, Canada;Bloorview Research Institute, Bloorview Kids Rehab and the Institute of Biomaterials and Biomedical Engineering, University of Toronto, Toronto, ON, Canada
Venue:
IEEE Transactions on Signal Processing
Year:
2010

Citing 3
Cited 1

A study of retrospective and on-line event detection

Proceedings of the 21st annual international ACM SIGIR conference on Research and development in information retrieval
Information Retrieval

Information Retrieval
Drowsy driver detection through facial movement analysis

HCI'07 Proceedings of the 2007 IEEE international conference on Human-computer interaction

A mobile food intake monitoring system based on breathing signal analysis

BodyNets '13 Proceedings of the 8th International Conference on Body Area Networks

Quantified Score

Hi-index	35.68

Visualization

Abstract

Swallow accelerometry is an emerging tool for non-invasive dysphagia screening. However, the automatic detection of a swallowing event is challenging due to contaminant vibrations arising from head motion, speech and coughing. In this paper, we consider the acceleration signal as a stochastic diffusion where movement is associated with drift and swallowing with volatility. Using this model, we develop a volatility-based swallow event detector that operates on the raw acceleration signal in an online fashion. With data from healthy participants and patients with dysphagia, the proposed detector achieves performance comparable to previously proposed swallow segmentation algorithms,with the added benefit of online detection and no signal pre-processing. The volatility-based detector may be useful for event identification in other biomechanical applications that rely on accelerometry signals.