This study aims to determine whether it can distinguish odontogenic keratocyst (OKC) and simple bone cyst (SBC) based solely on preoperative panoramic radiographs through a deep learning algorithm. (1) Methods: We conducted a retrospective analysis of patient data from January 2018 to December 2022 at Pusan National University Dental Hospital. This study included 63 cases of OKC confirmed by histological examination after surgical excision and 125 cases of SBC that underwent surgical curettage. All panoramic radiographs were obtained utilizing the Proline XC system (Planmeca Co., Helsinki, Finland), which already had diagnostic data on them. The panoramic images were cut into 299 × 299 cropped sizes and divided into 80% training and 20% validation data sets for 5-fold cross-validation. Inception-ResNet-V2 system was adopted to train for OKC and SBC discrimination. (2) Results: The classification network for diagnostic performance evaluation achieved 0.829 accuracy, 0.800 precision, 0.615 recall, and a 0.695 F1 score. (4) Conclusions: The deep learning algorithm demonstrated notable accuracy in distinguishing OKC from SBC, facilitated by CAM visualization. This progress is expected to become an essential resource for clinicians, improving diagnostic and treatment outcomes.

Objective: This study aimed to evaluate and compare the diagnostic performance of computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography-computed tomography (PET-CT) in detecting metastatic cervical lymph nodes in patients with oral squamous cell carcinoma (OSCC), with a particular focus on radiologically indeterminate lymph nodes. Materials and Methods: A retrospective analysis was conducted on OSCC patients who underwent CT, MRI, and PET-CT imaging prior to surgery, followed by histopathologic confirmation. Lymph nodes were categorized as metastatic, indeterminate, or benign based on imaging criteria specific to each modality. Diagnostic accuracy, sensitivity, and specificity were calculated using histopathology as the reference standard. Results: After excluding lymph nodes classified as indeterminate on preoperative imaging, CT demonstrated an accuracy of 83.6%, sensitivity of 51.9%, and specificity of 92.4% for metastatic lymph node detection. MRI and PET-CT showed comparable accuracies (81.6% and 80.8%, respectively) and identical sensitivities (53.9%). In contrast, among radiologically indeterminate lymph nodes, most lesions identified on CT and MRI were histopathologically benign, whereas PET-CT yielded very few indeterminate nodes, all of which were benign. For benign node identification, PET-CT exhibited the highest sensitivity (86.8%) but lower specificity (55.8%), while CT and MRI demonstrated more balanced specificity (73.1%) with lower sensitivity. Logistic regression confirmed SUVmax as a significant predictor of malignancy (p < 0.001; odds ratio 1.71, 95% CI: 1.48–2.35), and ROC analysis demonstrated strong discriminative performance (AUC = 0.88), with an optimal SUVmax cutoff of 3.6. Conclusion: While CT remains highly specific, PET-CT offers greater sensitivity in detecting benign and indeterminate lymph nodes, making it a valuable adjunct in preoperative assessment. SUVmax serves as a strong quantitative indicator for metastatic involvement. A multimodal imaging approach may enhance diagnostic accuracy, particularly in cases where lymph nodes lack definitive radiologic features.

Sungyeop Lee, Sung-Yun Hwang, Kyu-Bum Kim, Suyi Sim, Jihye Ryu, Chiho Moon, Jae-Yeol Lee, Dae Seok Hwang, Yong-Deok Kim, Sang-Hun Shin, Uk-Kyu Kim. JKAOMS 2026;52:187-95. https://doi.org/10.5125/jkaoms.2026.52.4.187

Jea-Young Ahn, Jung-Un Kim, Ji-Cheul Bae, Su-Yi Sim, You Jin Lee, Na Rae Choi. JKAOMS 2026;52:200-7. https://doi.org/10.5125/jkaoms.2026.52.4.200