AUTHOR=Gao Peimao , Huang Guowu , Zhao Lu , Ma Sen 

TITLE=Identification of biological indicators for human exposure toxicology in smart cities based on public health data and deep learning

JOURNAL=Frontiers in Public Health

VOLUME=Volume 12 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2024.1361901

DOI=10.3389/fpubh.2024.1361901

ISSN=2296-2565

ABSTRACT=With the acceleration of urbanization, the risk of urban populations being exposed to environmental pollutants is increasing. Safeguarding public health becomes paramount in the construction of smart cities. This research aims to propose a method for identifying biological indicators for urban human exposure toxicology in smart cities based on public health data and deep learning, aiming to achieve precise assessment and management of exposure risks. Initially, the study utilizes sensor networks within smart city infrastructures to collect environmental monitoring data, encompassing indicators such as air quality, water quality, and soil contamination. These data can be transmitted in real-time to databases within health departments, providing the groundwork for subsequent analysis. Leveraging public health data, a database containing information on types and concentrations of environmental pollutants is established. Associating this data with public health data supports subsequent research. Deep learning algorithms are employed to process and analyze environmental and health data. Firstly, convolutional neural networks are utilized for pattern recognition of environmental monitoring data, identifying relationships between different indicators and constructing a model correlating health indicators with environmental ones. Through training optimization, biological indicators associated with environmental pollution exposure are identified. Experimental analysis reveals that the predictive accuracy of the constructed model reaches 93.45%, thereby providing decision support for governments and health departments.