E-Cigarette Aerosol Characterization Using Holography and Machine Learning
Abstracts
Initial Award Abstract |
|
Electronic cigarettes (E-Cigs) are a growing health concern. There are hundreds of E-liquids with different compositions, additives, and flavorings - conclusive investigation of E-Cigs is very difficult, time consuming and costly. To better analyze the health implications of E-cigs to users and the public, there needs to be a simple, cost-effective, accurate and rapid analysis tool to characterize the resulting aerosol from E-cigs in field-settings. This proposal aims to address this need, with the following aims:
Aim 1: We will develop an enhanced beta-prototype of a field-portable aerosol analysis tools to characterize particles from E-cig emissions using computational imaging and deep learning. Using a cost-effective, wireless enabled device powered by holography and machine learning, we will quantify particle concentration, size distribution, and particle morphology in E-cig aerosol. This new particle imaging technology is fundamentally different compared to existing particle analyzers and will have a significantly larger dynamic range in terms of both particle size and density, as well as the ability to detect, classify and quantify volatile particles.
Aim 2: We will validate and characterize primary E-cig emissions (ECE)s immediately after generation and second hand vapors post-inhalation from E-cig users to determine the change in particle composition, post-inhalation. Together with Aim 1, this study will reveal a complete picture regarding the contents of the inhaled particles and the second hand vapors resulting from different E-cig devices and E-liquids.
Aim 3: We will determine the air quality in and around vape shops to identify potential risks to public health. Our results will provide evidence and new insights for policy makers, public health experts, as well as healthcare providers.
In addition to creating an innovative cost-effective and field-portable device with automated quantification and classification for large longitudinal study, the results from this proposed work will, for the first time, provide a complete end-to-end understanding of the evolution of E-cig, and the potentially harmful implications from primary ECE generation to long-term persistence in public environments. |
