Improving the Use of Sedative Medications in Dental Treatment of Children with Autism Spectrum Disorders Based on AI Technologies

Dental procedures are challenging for patients with autism spectrum disorder (ASD) due to behavioral disorders, high levels of fear, and sensory sensitivity. In order to ensure safe and effective treatment, various sedative medications are used in dentistry. This study investigated the potential of using artificial intelligence technologies to improve the use of appropriate sedative medications for effective dental procedures in children with ASD. Based on artificial intelligence technologies, Random Forest (RF), XGBoost, and Support Vector Machine (SVM) multi-class classifiers were used to select sedative medications for ASD patients and their accuracies were compared. The study examined the sedation classes of nitrous oxide inhalation, intranasal sedation, intravenous (IV) sedation, and general anesthesia. The models were trained and evaluated based on clinical parameters such as patient age, body weight, ASD level, cooperation level, anxiety index, and dental procedure complexity. A local clinical dataset was developed using anonymized patient records that included these features. Before entering the dataset into the model, pre-processing steps such as filling in missing values, coding categorical columns, and balancing class imbalance were performed. According to the results, the XGBoost model demonstrated the highest overall performance with Accuracy = 96.8%, F1-score = 96.4%. This demonstrated the model’s high predictive ability in data with complex and unbalanced class distribution. The proposed AI approach can optimize the clinical decision-making process by allowing the safe and effective organization of dental treatment in ASD patients, allowing the selection of the type of sedation appropriate to the individual characteristics of the patient.

Ҳали таржима қилинмаган