Related Experiment Video
Updated: Jun 27, 2026
05:49
Reliability of Artificial Intelligence-Based Cone Beam Computed Tomography Integration with Digital Dental Images
Published on: February 23, 2024
Artificial Intelligence in Orofacial Pain: Diagnostic and Predictive Performance Across Machine Learning and Deep
Laura Iosif1, Marina Imre1, Andreea Gabriela Wagner1
1Department of Prosthodontics, Faculty of Dentistry, Carol Davila University of Medicine and Pharmacy, 010232 Bucharest, Romania.
Diagnostics (Basel, Switzerland)
|June 26, 2026
Summary
Artificial intelligence (AI) shows promise for diagnosing orofacial pain (OFP). AI performs best for neurovascular and musculoskeletal pain, but challenges remain for odontogenic and neuropathic pain diagnosis.
Area of Science:
- Dentistry and Oral Health
- Neurology
- Artificial Intelligence in Medicine
Background:
- Orofacial pain (OFP) encompasses diverse conditions with overlapping symptoms, complicating accurate diagnosis.
- Increasing diagnostic challenges drive interest in artificial intelligence (AI) for enhancing OFP diagnostic precision.
Purpose of the Study:
- To review the application of AI in the diagnosis, classification, and prediction of adult OFP.
- To assess the diagnostic performance of AI across different OFP categories.
Main Methods:
- A narrative review of studies published between 2016-2026 was conducted.
- Searches were performed in PubMed/MEDLINE, Scopus, and Web of Science.
- Included studies applied AI to OFP diagnosis and reported at least two performance metrics.
Main Results:
- AI demonstrated high and consistent diagnostic performance for neurovascular pain (e.g., migraine) and musculoskeletal pain (e.g., TMDs).
- Odontogenic pain diagnosis showed lower and more variable AI performance, with imaging-based models yielding better results.
- Neuropathic pain exhibited moderate to high performance in specific radiomics studies, but results were inconsistent due to variability.
Conclusions:
- AI holds significant potential for OFP diagnosis, particularly for neurovascular and musculoskeletal pain.
- Clinical translation is hindered by data heterogeneity and a lack of validation.
- Future progress requires multimodal datasets and multicenter studies for robust and generalizable AI tools.
