PREDICTION OF SILICOSIS IN IRON AND STEEL WORKERS USING ARTIFICIAL INTELLIGENCE

Abstract

Silicosis is one of the most common occupational lung diseases in Egypt, where its prevalence rate ranges from 18.5 % to 45.8% among workers exposed to free crystalline silica dust. Long-term contact to silica dust, which is prevalent in the iron and steel sector, can lead to silicosis, a crippling occupational lung disease. Periodic health tracking and worker medical evaluations are traditional methods for finding and managing the risk of silicosis; however, these techniques frequently have limitations in terms of accuracy and dependability. It is feasible to analyse sizable databases of worker health and exposure information using artificial intelligence. The prevalence and severity of silicosis in iron and steel workers may be lessened due to development of more precise and effective risk evaluation and management methods. This study aims to predict iron and steel workers liability to developing silicosis based on artificial intelligence. A case-control study design was used to fulfill the aim of this study. The study was conducted at the El Gedida Iron Mine area at Bahariya Oasis, Giza Governorate. A sample of 220 workers was included in this study. The participants received a questionnaire in addition to a standard medical examination. The researchers chose eight variables that influence silicosis based on a literature review then suggested an ensemble model based on the Feed Forward Neural Network, K-Nearest Neighbor, Fuzzy K-Nearest Neighbor, and Decision Tree. The experimental findings reveal that the suggested framework surpassed others in the automated prediction of silicosis, obtaining an accuracy of 97.2 percent. Keywords: Silicosis, K-Nearest Neighbor, Fuzzy K-Nearest Neighbor, Decision Tree DOI： https://doi.org/10.35741/issn.0258-2724.58.2.15