Using Artificial Intelligence to Detect Periapical Lesions on Dental Panoramic Images Compared to Cone Beam CT

What this Trial Is About

Researchers are investigating ways to improve the early detection of periapical lesions, which are dental damages visible as radiolucent areas on various imaging methods. These lesions can lead to serious local and systemic health problems, including infections and cardiovascular diseases. The study focuses on developing an artificial intelligence algorithm that can better identify these lesions on panoramic dental radiographs, which are commonly used but less sensitive than 3D imaging techniques like cone-beam computed tomography CBCT. The study involves using a deep learning system trained on reference data that includes both panoramic dental images and CBCT scans taken within three months of each other. This approach aims to enhance the accuracy of lesion detection compared to conventional radiographic methods. The AI model builds upon previous work and incorporates advanced techniques such as transfer learning to improve data quality and detection performance. Participants will have their existing panoramic and CBCT images analyzed by the AI software over a two-year period to assess its performance and specificity. The research team will compare how well the AI detects periapical lesions against the current gold standard imaging. This observational study does not involve new treatments but focuses on image analysis and diagnostic improvement to support faster and more effective care for affected patients.

Detection of Periapical Lesions on Dental Panoramic Radiographs Based on Artificial Intelligence