PosterPDF Available

Smartphone-based Cough and Sleep Quality Detection

Authors:
  • Resmonics AG

Abstract

We are investigating to which degree of accuracy can a mobile application detect asthmatic nocturnal cough and sleep quality with the smartphone’s built-in microphone?
www.c4dhi.org
Smartphone-based Cough and
Sleep Quality Detection
1. Problem 2. Research Question
To which degree of accuracy can a mobile
application detect asthmatic nocturnal cough
and sleep quality with the smartphone’s built-in
microphone?
1. Marsden et al. (2016): Nocturnal cough frequency
provides an objective assessment of asthma
symptoms that correlates with standard
measures of asthma control
2. Luyster et al. (2012): Sleep quality is associated
with asthma control even if accounted for
concomitant diseases
4. Method:Learning Pipeline
5. Expected Results
References
Barata, F., Kowatsch, T., Tinschert, P., Filler, A., Personal MobileCoach: TailoringBehavioral Inter ventions to the
Needs of Individual Participants, UbiComp ’16 Proceedings of the 2016 ACM International Joint Conference on
Pervasive and Ubiquitous Computing: Adjunct Workshop Designing, Developing, and Evaluating The Internet of
Personal Health (IoPH), Heidelberg, Germany, 1089-1094.
Luyster, Faith S., et al. "Sleep quality and asthma contro l and quality of life in non-severe and severe asthma." Sleep
and Breathing 16.4 (2012): 1129-1137.
Marsden, Paul A., et al. "Objective cough frequency, ai rway inflammation, and disease control in asthma." CHEST
Journal 149.6 (2016): 1460-1466.
Tinschert, P., Barata, F., Kowatsch, T., Enhancing Asthma Control through IT: Design, Impleme ntation and Planned
Evaluation of the Mobile Asthma Companion, in Leimeister, J.M.; Brenner, W. (Hrsg.): Proceedings de r 13th
International Conference on Wirtschaftsinformatik (WI 2017), St. Gallen, 1291-1294.
EPFL Lausanne | January 27-30 | 2018Applied Machine Learning Days
Partner
lower airways
"

"
paint
Triggers
Symptoms
chest
tightness
cough
expiratory
wheezing
shortness
of breath
sleep
quality
«cough»
«no cough»
Raw Sensor Data
Pre-processing
Spectrograms
Convolutional
Neural
Network
Training Evaluation
Leave-One-
Person-Out
Cross-
validation
Awake
Asleep
8.00 pm 10.00 am
Cough
A classification model with accuracy values close to 1 for performing
the cough detection and sleep quality estimation can be developed.
Filipe Barata
1
, Peter Tinschert
2
, Frank Rassouli
3
, Florent Baty
3
, Martin Brutsche
3
,
Claudia Steurer-Stey
4,5
, Milo Puhan
4
, Elgar Fleisch
1,2
& Tobias Kowatsch
2
1 ETH Zurich, 2 University of St.Gallen, 3Cantonal Hospital St.Gallen, 4 University of Zurich & 5 medix Zurich
3. Research Framework
Example: cough detection pipeline
ResearchGate has not been able to resolve any citations for this publication.
Conference Paper
Full-text available
The personal and financial burden of asthma highly depends on a patient's disease self-management skill. Scalable mHealth apps, designed to empower patients, have the potential to play a crucial role in asthma disease management. However, the actual clinical efficacy of mHealth asthma apps is poorly understood due to the lack of both methodologically sound research and accessible evidence-based apps. We therefore apply design science with the goal to design, implement and evaluate a mHealth app for people with asthma, the Mobile Asthma Companion (MAC). The current prototype of MAC delivers health literacy knowledge triggered by nocturnal cough rates. We conclude by proposing a randomized controlled trial to test the efficacy of our prototype.
Conference Paper
Full-text available
MobileCoach, an open source behavioral intervention platform, has been developed to provide health professionals with an authoring tool to design evidence-based, scalable and low-cost digital health interventions (DHI). Its potential meets the lack in resources and capacity of health care systems to provide DHI for the treatment of noncommunicable diseases. In the current work, we introduce the first personalization approach for MobileCoach with the purpose of identifying the needs of participants, tailoring the treatment and, as a consequence, enhancing the capability of MobileCoach-based DHIs. The personalization approach is then exemplified by a very first prototype of a DHI for people with asthma that is able to detect coughing by just using a smartphone's microphone. First empirical results with five healthy subjects and 80 coughs indicate its technical feasibility as the detection accuracy yielded 83.3%. Future work will focus on the integration of personalized sensing and supporting applications for MobileCoach.
Article
Full-text available
Background: Cough is recognised as an important troublesome symptom in the diagnosis and monitoring of asthma. Asthma control is thought to be determined by the degree of airway inflammation and hyper-responsiveness but how these relate to cough frequency is unclear. Objective: To investigate the relationships between objective cough frequency, disease control, airflow obstruction and airway inflammation in asthma. Methods: Participants with asthma underwent 24 hour ambulatory cough monitoring, exhaled nitric oxide, spirometry, methacholine challenge and sputum induction (cell counts and inflammatory mediator levels). Asthma control was assessed by GINA classification and the Asthma Control Questionnaire (ACQ). Results: Eighty-nine subjects with asthma (mean age 57 years (±SD 12); 57% female) were recruited. According to GINA criteria, 18 (20.2%) patients were classified as controlled, 39 (43.8%) partly controlled and 32 (36%) uncontrolled; median (range) ACQ score was 1 (0.0-4.4). ACQ-6 correlated with 24hr cough frequency (r=0.40; p<0.001) and patients with uncontrolled asthma (GINA) had higher median 24hr cough frequency (4.2c/h, range 0.3-27.6) compared with partially controlled and controlled asthma (1.8c/h, range 0.2-25.3 and 1.7c/h range 0.3-6.7, p=0.01 and p=0.002 respectively). Measures of airway inflammation were not significantly different between GINA categories and were not correlated with ACQ. In multivariate analyses, increasing cough frequency and worsening FEV1 independently predicted measures of asthma control. Conclusion: Ambulatory cough frequency monitoring provides an objective assessment of asthma symptoms that correlates with standard measures of asthma control, but not airflow obstruction or airway inflammation. Moreover, cough frequency and airflow obstruction represent independent dimensions of asthma control.
Article
Full-text available
Purpose The effect of sleep quality on asthma control independent from common comorbidities like gastroesophageal reflux disease (GERD) and obstructive sleep apnea (OSA) is unknown. This study examined the association between sleep quality and asthma control and quality of life after accounting for OSA and GERD in non-severe (NSA) and severe (SA) asthma. Methods Cross-sectional data from 60 normal controls, 143 with NSA, and 79 with SA participating in the Severe Asthma Research Program was examined. Those who reported using positive airway pressure therapy or were at high risk for OSA were excluded. Results Both SA and NSA had poorer sleep quality than controls, with SA reporting the worst sleep quality. All asthmatics with GERD and 92% of those without GERD had poor sleep quality (p = 0.02). The majority (88–100%) of NSA and SA participants who did not report nighttime asthma disturbances still reported having poor sleep quality. In both NSA and SA, poor sleep quality was associated with worse asthma control and quality of life after controlling for GERD and other covariates. Conclusions These results suggest that poor sleep quality is associated with poor asthma control and quality of life among asthmatics and cannot be explained by comorbid GERD and nighttime asthma disturbances.