Association between tooth agenesis and developmental dental anomalies: an observational study with machine learning analysis.

OBJECTIVE: To investigate the prevalence of tooth agenesis and its association with other developmental dental anomalies in non-syndromic patients. The study also evaluated the applicability of machine learning models to predict the occurrence of agenesis. METHODS: This cross-sectional observational study analyzed 4,990 panoramic radiographs from radiology centers in Uberlândia, Minas Gerais/Brazil. It assessed the presence of tooth agenesis and associated anomalies, such as palatal displacement of canines, distoangulation of lower second premolars, tooth transposition, infraocclusion of deciduous molars, mesioangulation of lower second molars, and supernumerary teeth. The tooth agenesis code (TAC) mapped the patterns of occurrence. Fisher's exact test, Student's t-test, and logistic regression were used to analyze the data (p < 0.05). The machine learning models used included Decision Tree, Random Forest, XGBoost, Support Vector Machine, and Deep Learning. RESULTS: Agenesis was found in 6.1% of patients, with a higher prevalence in females (57.6%) and younger individuals (mean age of 16.8 years). Dental transposition (OR: 19.11; p < 0.001), maxillary canine displacement (OR: 2.08; p < 0.001), and deciduous molar infraocclusion (OR: 147.73; p < 0.001) showed a significant association. The Random Forest model showed the best predictive performance, but with clinical limitations. CONCLUSION: Tooth agenesis is associated with an increased risk of other anomalies, highlighting the importance of early diagnosis and integrated therapeutic approaches.