Prediction of premolar tooth lengths based on their panoramic radiographic lengths

OBJECTIVE: Establishing a linear regression model for the determination of first premolar lengths based on panoramic radiographs. MATERIALS AND METHODS: The real lengths of 112 first premolars extracted for orthodontic reasons were measured and compared with their panoramic lengths. The teeth were divided into four groups according to their intraoral quadrant locations (T14, T24, T34, T44) and regression analysis was conducted for each group. RESULTS: A linear regression model for the prediction of tooth length (mm) based on the panoramic length was established (P < 0.0001): For group T14 the predicted length = (panoramic length x 0.698) + 2.61. For group T24 the predicted length = (panoramic length x 0.5056) + 7.844. For group T34 the predicted length = (panoramic length x 0.5075) + 9.282. For group T44 the predicted length = (panoramic length x 0.436) + 11.298. CONCLUSION: Prediction of all first premolar lengths using their panoramic images is both feasible and reliable.