Development of an application for estimating the personal exposure to air particulate matter using wearable sensors

One of the risk factors that increases the asthma patient’s risk of exacerbation is exposure to air pollutants, in particular a short-term exposure to particulate matter has been associated with asthma exacerbations and hospital visits. Based on that, the exposure assessment of asthma patients, meaning the process of measuring or estimating the intensity of exposure, is a topic of great interest. This can be done by using fixed air quality monitoring stations and obtaining mean results in terms of space and time, but this solution is limiting since it does not consider patient’s movement and it does not make differences between indoor and outdoor exposure. Subjects spend in fact most of their time indoor, at home or at workplace, and it is important to consider also indoor air pollution. To overcome these issues, personal exposure can be assessed using wearable/portable sensors. In this thesis, we propose a new approach to monitor the personal exposure to particulate matter using a portable air quality sensor, the Atmotube PRO by Atmo (San Francisco, CA, USA), a wearable activity tracker by Fitbit Inc. (San Francisco, CA, USA), and a mobile application for data integration and analysis. In particular, the application integrates four literature models that allow to estimate the minute ventilation (i.e. the volume of inhaled air per minute) starting from heart rate measurements fetched from the Fitbit sensor. Combining the minute ventilation with the timeseries of particulate matter concentration measure by the Atmotube PRO sensore, the app then estimates the inhaled dose of these pollutants over the monitoring period. The application developed in this thesis using Flutter framework and Dart language exploits the Fitbit Web API and the the Atmotube Bluetooth API to fetch data from the Fitbit server and from the Atmotube PRO device respectively. Two packages based on the APIs, fitbitter and atmotuber, were used to authorize Fitbit and connect Atmotube PRO device and to make requests for data. The database was created using the drift package and the app state was managed with the bloc package, while the graphics were made using fl_chart. The developed application will allow the patients with asthma to continuously monitor: i) air quality indices, in particular the concentrations of PM1, PM2.5, PM10 and VOC, ambient temperature, humidity, and pressure; ii) activity data, specifically the heart rate and steps; ii) model-based estimates of minute ventilation and inhaled pollutant dose. Timeseries of inhaled pollutant dose are further processed to extract relevant exposure indices, such as the total inhalation of particulate matter for each day. In the last part of the thesis, the Atmotube and Fitbit data of a representative subject are analysed in details to evaluate the collected signals visually, compute relevant statistical parameters, and compare the estimates of minute ventilation and inhaled pollutant dose obtained by the four implemented models.