Objective: To explore and evaluate the feasibility and accuracy of using optical coherence tomography and optical coherence microscope (OCT/OCM) for diagnosis of oral cancer. Methods: In this study, OCT/OCM was utilized to image the oral mucosa specimens. A total of 289 ex vivo oral mucosa specimens were collected from 68 patients with oral cancer who were hospitalized at Department of Oral and Maxillofacial Head and Neck Tumors, Capital Medical University School of Stomatology, between January 2021 and February 2023, resulting in a dataset of 1 445 OCT/OCM images. By observing the characteristic patterns in the OCT/OCM images, including normal oral mucosa, epithelial abnormal proliferation (mild, moderate, severe), and oral cancer, these patterns were matched with corresponding pathological images. A diagnostic study was conducted, employing pathological diagnosis as the gold standard and utilizing a double-blind experimental design involving three diagnostic evaluators who participated in the analysis and diagnosis of OCT/OCM images. Results: The OCT/OCM images demonstrated good correlation with the corresponding pathological images, and diagnostic criteria were established based on the comparative results. In the diagnostic study involving three investigators, the accuracy was 82%, sensitivity was 84% (95%CI: 80%-88%), and specificity was 81% (95%CI: 77%-85%). There was a high level of agreement among the observers (κ=0.614), indicating substantial concordance in the diagnostic results among the three investigators. Conclusions: This study demonstrates the potential of OCT/OCM for diagnosis of oral cancer. The technology accurately distinguishes between normal oral mucosa, epithelial abnormal proliferation and oral cancer.
Read full abstract