Prediction of Asthma Attacks Using ANFIS and Mobile Technologies

The COVID-19 pandemic has significantly accelerated the adoption of computer and communication technologies in disease diagnosis and healthcare, particularly in the study of respiratory conditions. Diagnosis of many respiratory diseases, including asthma, often relies on assessing indoor and outdoor air quality around the patient, along with specific tests related to the patient’s condition. This study aims to predict asthma attacks using Adaptive Neuro-Fuzzy Inference System (ANFIS) technology based on Air Quality Index (AQI) and Peak Expiratory Flow Rate (PEFR) data. The system integrates an embedded microcontroller and IoT sensor technology to monitor levels of crucial gases such as ammonia, hydrogen sulfide, carbon dioxide, acetone, formaldehyde, humidity, temperature as well as PM2.5, PM10. This comprehensive approach utilizes both AQI and PEFR data to predict asthma attacks, issue alerts, offer medical advice through a mobile application to asthma patients, and transmit patient conditions and related data to medical centers in critical cases. Test results indicate high accuracy in predicting both air quality index and asthma severity, with minimal errors of 0.05 RMSE for AQI and 0.0045 RMSE for asthma severity.