Intelligent Asthma Self-management System for Personalised Weather-Based Healthcare Using Machine Learning

• Main Authors: Haque, R., Ho, S.-B., Chai, I., Teoh, C.-W., Abdullah, Adina, Tan, C.-H., Dollmat, K.S.
• Format: Article
• Published: Springer Science and Business Media Deutschland GmbH 2021
• Subjects: R Medicine; RA Public aspects of medicine; RA0421 Public health. Hygiene. Preventive Medicine
• Online Access: http://eprints.um.edu.my/35801/; https://www.scopus.com/inward/record.uri?eid=2-s2.0-85112711727&doi=10.1007%2f978-3-030-79457-6_26&partnerID=40&md5=7e46d3ae4f8ee429b5bdee6875377309

Asthma is a common chronic disease that affects people from all age groups around the world. Although asthma cannot be cured, strategies to enhance applications on self-management can be effective to control asthma exacerbations. In recent years, researchers have been developing various mHealth tools and applications for self-management. However, there is a lack of effective personalised self-management solution for asthma that can be adopted widely. Personalisation is important for identifying each patient’s demographic characteristics, measuring their asthma severity level, and most importantly, predicting the triggers of asthma attacks. It has been observed that weather attributes (e.g. temperature, humidity, air pressure and thunderstorms) impact on triggering asthma attacks and adversely affect the symptoms of asthmatic patients. Hence, developing an intelligent asthma self-management system for personalised weather-based healthcare using machine learning technique can help predict weather impact on asthma exacerbations for individual patients and provide real-time feedback based on daily weather forecasts. Therefore, this paper explores the impact of weather on asthma exacerbations and examines the effectiveness and limitations of several recent asthma self-management tools and applications. Consequently, based on the uses and gratifications theory, an engineering model for personalised weather-based healthcare is proposed which incorporates major constructs including mHealth application, asthma control test, demographic characteristics, weather attributes, machine learning technique and neural networks. © 2021, Springer Nature Switzerland AG.