Developing an Asthma Attack Alert System by Signal Processing of Breathing Sounds

Objective: Asthma is a widespread cause of hospitalizations and even deaths across all age groups worldwide, making it a significant health concern. Its prevalence and burden have been reported in Turkey, with allergic rhinitis frequencies ranging from 11.8% to 36.4% in five different centers. Asthma prevalence in Turkey varies between 6% and 15% in children and between 2% and 17% in adults. In this context, the aim is to design and develop a device capable of detecting asthma attacks by analyzing respiratory sounds and alerting the patient. The project's objectives include providing rapid pre-attack warnings without disturbing the patient, employing an air pollution sensor, and incorporating two distinct warning mechanisms. Methods: This research employs a series of methods to analyze the respiratory sounds of asthma patients and detect asthma attacks. Initially, a literature review was conducted to identify critical factors for detecting asthma attacks, focusing on developing new technologies and non-invasive methods to obtain respiratory sounds. Arduino and MATLAB tools were employed to develop sound analysis and data processing. Subsequently, filtering and statistical calculations were performed. In the final stage, various visualizations such as sound signal, filtered sound signal, FFT, and power spectrum were created to present the obtained results through a user-friendly interface. The two-stage alert system incorporates the ability to detect asthma attacks based on wheezing sounds and air pollution. Results: In consideration of the prevalence and severity of asthma, this study addresses a device designed for early diagnosis of the attack. The device is equipped to analyze respiratory sounds within the frequency range of 100-1000 Hz, enabling differentiation between wheezes and rhonchi. The utilization of this device has the potential to enhance the quality of life for asthma patients and provide the opportunity for early intervention by anticipating attack situations. Conclusion: By focusing on the analysis of respiratory sounds, this research demonstrates the feasibility of monitoring asthma attacks through non-invasive methods. The developed programs and equipment have the potential to offer asthma patients a more effective way of life.