Distribution of high fluoride in groundwater and development of GIS-based decision support system in health information systems for dental fluorosis

Fluoride is a crucial element for the formation and maintenance of bones and teeth. When the optimal daily dose of fluoride is exceeded by various means, it can cause a range of systemic effects in the body, depending on the amount of the dose exceeded and the duration of intake. When fluoride ions are taken at the optimal level, they demonstrate an anticariogenic effect on teeth. The health effects resulting from long-term ingestion of fluoride include dental and skeletal fluorosis, which in most cases has a significant impact on the lives of individuals. In recent years, the diversity of potential fluoride sources has led to an increase in the prevalence of dental fluorosis. Consequently, the effects of fluoride, the mechanisms by which it acts, the sources of fluoride intake, the duration of intake and the diagnosis of fluorosis continue to be the subject of extensive research. It is also important to note that endogenous and/or exogenous factors unrelated to fluoride exposure can also cause dental enamel staining. The project team aimed to conduct a comprehensive assessment of the number of children (aged 6-15 years) potentially exposed to fluoride in the study area and the associated fluoride hazard and risk. In this context, the primary objective of the study was to ascertain the prevalence of dental fluorosis in the Sanliurfa province, which is a semi-arid region, and to elucidate the relationship between the fluoride content of drinking water and the occurrence of fluorosis. In the current project, water samples were collected from various locations within Sanliurfa province and were subsequently analyzed in conjunction with dental examinations. The study involved the implementation of dental fluorosis screenings in 40 villages and neighborhoods, as well as 30 schools.The fluoride levels in the study area were mapped using Geographic Information Systems (GIS). A total of 45 water samples were collected from the regions where cases were detected and analyzed. Fluoride concentration levels in the study area's water resources exhibit considerable variability, with annual averages ranging from 0.26 to 5.62 mg/L. Furthermore, a non-carcinogenic risk assessment was conducted for two age groups. The order of health risks in the study groups was as follows: children > adolescents. The average hazard coefficient (HQ) value was calculated in the range of 0.139495 to 7.184253 for the two age groups. In particular, the study found that schoolchildren in this region exhibited severe deterioration and staining of their teeth, with some cases of fourth and fifth degree fluorosis.In the final phase of the project, dental images were analyzed using artificial intelligence and deep learning techniques. The dataset, comprising a total of 76 subjects, was divided into three subsets for the purpose of training the Yolov8 object detection model: a training set, a validation set and a test set. The separation process was subjected to the shuffling process by assigning a random seed value of 123 to the file list. Following approximately 100 epochs, the model demonstrated a plateau in performance and loss value, indicating that further training at higher epochs was unnecessary. The results of the model training on dental images indicated successful classification of the degree of fluorosis.