Accuracy of panoramic radiography, periapical radiography, and CBCT in diagnosing apical root resorption

This study evaluated the accuracy of digital panoramic radiography, digital periapical radiography, and cone-beam computed tomography (CBCT) in diagnosing apical root resorption (ARR) using 85 extracted teeth from dry human skulls and mandibles. Four ARR levels were simulated: mild, moderate, severe, and extreme. CBCT scans, panoramic radiographs, and periapical radiographs were obtained before and after ARR simulation. Three examiners independently evaluated all images to determine ARR and, when present, to classify its severity level. Inter- and intra-observer agreement was calculated using Cohen's kappa, and sensitivity, specificity, and accuracy were determined for each method. Receiver operating characteristic (ROC) curve analysis was performed to obtain the area under the curve (AUC). AUCs were compared using the DeLong test. All imaging methods showed substantial inter- and intra-examiner agreement (κ > 0.60). CBCT demonstrated the highest sensitivity, specificity, and accuracy, followed by periapical and panoramic radiography. CBCT was significantly superior (p < 0.01) in detecting mild (AUC = 0.939) and moderate (AUC = 0.947) ARR compared with the other methods (AUC < 0.877). For severe ARR, CBCT (AUC = 0.953) and periapical radiography (AUC = 0.921) performed similarly (p > 0.05) and were superior (p=0.03) to panoramic radiography (AUC = 0.874). For extreme ARR, no significant difference was observed among the methods. In conclusion, CBCT showed high accuracy in detecting ARR, particularly at milder levels (< 2 mm). Periapical radiography was comparable to CBCT for severe resorption (one-third of the root), while all methods performed similarly for extreme resorption (half of the root). Thus, CBCT appears most suitable for early ARR detection, whereas periapical and panoramic radiography are adequate for diagnosing advanced stages.