Unorganized Machines to Estimate the Number of Hospital Admissions Due to Respiratory Diseases Caused by PM10 Concentration

Abstract

The particulate matter PM10 concentrations have been impacting hospital admissions due to respiratory diseases. The air pollution studies seek to understand how this pollutant affects the health system. Since prediction involves several variables, any disparity causes a disturbance in the overall system, increasing the difficulty of the models’ development. Due to the complex nonlinear behavior of the problem and their influencing factors, Artificial Neural Networks are attractive approaches for solving estimations problems. This paper explores two neural network architectures denoted unorganized machines: the echo state networks and the extreme learning machines. Beyond the standard forms, models variations are also proposed: the regularization parameter (RP) to increase the generalization capability, and the Volterra filter to explore nonlinear patterns of the hidden layers. To evaluate the proposed models’ performance for the hospital admissions estimation by respiratory diseases, three cities of São Paulo state, Brazil: Cubatão, Campinas and São Paulo, are investigated. Numerical results show the standard models’ superior performance for most scenarios. Nevertheless, considering divergent intensity in hospital admissions, the RP models present the best results in terms of data dispersion. Finally, an overall analysis highlights the models’ efficiency to assist the hospital admissions management during high air pollution episodes.

A Multi-Pollutant and Meteorological Analysis of Cardiorespiratory Mortality among the Elderly in São Paulo, Brazil—An Artificial Neural Networks Approach

“…Some studies, for instance, considered only pollutant concentrations [ 75 , 76 ]. Other studies included meteorological variables, days of the week, and traffic data as explanatory variables as well [ 77 , 78 ]. The presentation of the model’s quality also varies among the studies, with indicators such as mean square error (MSE), mean error (ME), mean absolute percentage error (MAPE), and R-square (R 2 ).…”

“…They tested lags, and the best-fitted ANN reached a MAPE equal to 26%. Tadano et al [ 78 ] considered the same input and output variables and obtained an ANN with a MAPE of 35%. Finally, Kachba et al [ 77 ] considered CO, NO x , O 3 , SO 2 , and PM concentrations and traffic data to infer respiratory hospital admissions and mortality and obtained a MAPE of 28% when analyzing the hospital admissions and 34% when analyzing the mortality.…”

Ambient PM Concentrations as a Precursor of Emergency Visits for Respiratory Complaints: Roles of Deep Learning and Multi-Point Real-Time Monitoring

“…Given the advancement of machine learning algorithms and processes, recent works have contributed to this literature by examining how deep learning models can be used to improve predictions for diseases triggered by ambient air pollution [10][11][12][13]. These learning techniques are important for the accurate prediction of hospital admissions or emergency room visits as they could help optimize the allocation of scarce medical resources.…”

“…Recently, Lu et al [10] used three different machine learning models to predict emergency room visits in Beijing, China, showing that the LSTM model reaches the highest accuracy with an advantage in the detection of a lag effect. Tadano et al [11] used two classic ANN architectures, extreme learning machines (ELM) and echo state networks (ESN), to expand the literature on ML applications for predicting respiratory-related hospital admissions due to PM 10 concentrations, indicating that the ELM model was the best predictor of daily hospital admissions for respiratory diseases given the PM 10 concentration, relative humidity, and ambient temperature inputs.…”

“…They tend to take air pollution measurements from one or only a few monitoring stations to calculate the daily averages of various pollutants, assuming all residents within the region are exposed to the same level of ambient air pollutants. However, measurements vary between stations and could lead to errors or biases when averaging, and thus the process of predicting emergency room visits due to respiratory illnesses could be improved by incorporating multipoint air pollution data into an advanced learning algorithm [11].…”