Detecting Episodes of Increased Cough Using Kinetic Earables

This paper introduces the detection of episodes of increased cough (e.g. during illness) based on cough event classification using kinetic earables. In a twelve subject study, we collected voluntary weak and strong cough as well as five non-cough activities (e.g., talking) under various conditions (e.g., walking). During the activities, an in-ear worn sensor records acceleration and gyroscope data. In total, we collected 4,200 activity samples. A single step classification pipeline (0.77 overall accuracy) serves as the foundation for statistical analysis to achieve episodes of increased cough discrimination. As a digression, we reverse data and perform pose classification which could enable faster cough episode prediction. All-in-all, earables might help to objectify illness to encourage formal diagnosis.

[1]  Matthias Budde,et al.  Towards Respiration Rate Monitoring Using an In-Ear Headphone Inertial Measurement Unit , 2019, EarComp@UbiComp.

[2]  Sozo Inoue,et al.  Human activity recognition using earable device , 2019, UbiComp/ISWC Adjunct.

[3]  George R. Wodicka,et al.  Detection of respiratory sounds at the external ear , 2004, IEEE Transactions on Biomedical Engineering.

[4]  Jakob E. Bardram The CAMS eSense Framework: Enabling Earable Computing for mHealth Apps and Digital Phenotyping , 2019, EarComp@UbiComp.

[5]  Thierry Dutoit,et al.  Objective Study of Sensor Relevance for Automatic Cough Detection , 2013, IEEE Journal of Biomedical and Health Informatics.

[6]  Chulhong Min,et al.  Earables for Personal-Scale Behavior Analytics , 2018, IEEE Pervasive Computing.

[7]  K G Moons,et al.  Clinical signs and symptoms predicting influenza infection. , 2001, Archives of internal medicine.

[8]  Pablo Casaseca-de-la-Higuera,et al.  Robust Detection of Audio-Cough Events Using Local Hu Moments , 2019, IEEE Journal of Biomedical and Health Informatics.

[9]  G. Dinant,et al.  The predictive value of influenza symptomatology in elderly people. , 1998, Family practice.

[10]  Marian Haescher,et al.  aHead: Considering the Head Position in a Multi-sensory Setup of Wearables to Recognize Everyday Activities with Intelligent Sensor Fusions , 2015, HCI.

[11]  K. Chung,et al.  Coughing frequency in patients with persistent cough: assessment using a 24 hour ambulatory recorder. , 1994, The European respiratory journal.

[12]  P. Klepac,et al.  Feasibility of controlling COVID-19 outbreaks by isolation of cases and contacts , 2020, The Lancet Global Health.

[13]  Paul Lukowicz,et al.  In the blink of an eye: combining head motion and eye blink frequency for activity recognition with Google Glass , 2014, AH.

[14]  J. Smith,et al.  Monitoring chronic cough: current and future techniques , 2010, Expert review of respiratory medicine.

[15]  Andreas W. Kempa-Liehr,et al.  Time Series FeatuRe Extraction on basis of Scalable Hypothesis tests (tsfresh - A Python package) , 2018, Neurocomputing.

[16]  F. Kawsar,et al.  Augmenting Conversational Agents with Ambient Acoustic Contexts , 2020, MobileHCI.

[17]  Merryn J Mathie,et al.  Accelerometry: providing an integrated, practical method for long-term, ambulatory monitoring of human movement , 2004, Physiological measurement.

[18]  Timothy J Cunningham,et al.  Trends in Self-Reported Sleep Duration among US Adults from 1985 to 2012. , 2015, Sleep.

[19]  Tianqi Chen,et al.  XGBoost: A Scalable Tree Boosting System , 2016, KDD.

[20]  Brian Subirana,et al.  COVID-19 Artificial Intelligence Diagnosis Using Only Cough Recordings , 2020, IEEE Open Journal of Engineering in Medicine and Biology.