Artificial Intelligence Detection of Effluent-Induced Gill Damage in River Fish

Abstract

Industrial effluent discharge is becoming a common issue in the ecosystem of rivers that releases toxic substances that change the environmental condition of the rivers and affect fish health. Fish gills are the most sensitive organs that react to pollution because they come into direct contact with the water, and effluent exposure may lead to abnormality in the structure like epithelial injury, lamellar injury, and tissue degeneration. But the current method of histopathological evaluation of gill damage needs expertise and consumes a lot of time. In this study, a new approach using artificial intelligence is suggested to detect effluent exposure in the gills of river fish through environmental and biological factors. Water quality factors such as pH level, dissolved oxygen, chemical oxygen demand, turbidity, and concentration of pollutants were investigated along with gill tissue analysis. Fish specimens were divided into healthy, moderate injury, and severe injury groups according to the histology analysis. Machine learning and deep learning approaches were used to discover the relationship between pollution factors and gill injury. The study found that increasing effluent exposure caused more injury to the fish gills, whereas the proposed AI approach succeeded in improving the prediction accuracy with 96.4%, precision with 96.0%, recall with 95.8%, and F1-Score with 95.9%. The findings revealed that artificial intelligence can be effectively used for fast evaluation of pollution effects on fish health through analyzing environmental and biological factors. The proposed model can serve as a valuable tool for aquatic ecosystems monitoring and river pollution control, and its future application may involve automatic environmental monitoring and wildlife health assessment