Artificial intelligence to support early diagnosis of temporomandibular disorders: A preliminary case study. (3rd November 2022)

• Record Type: Journal Article
• Authors: Reda, Bachar; Contardo, Luca; Prenassi, Marco; Guerra, Enrico; Derchi, Giacomo; Marceglia, Sara
• Is Part Of: Journal of oral rehabilitation. Volume 50:Number 1(2023)
• Page Start: 31
• Page End: 38
• Publication Date: 2022-11-03
• Subjects: artificial intelligence -- cognitive computing -- decision support system -- early diagnosis -- temporomandibular disorders; Dentistry -- Periodicals; Prosthodontics -- Periodicals; 617
• DOI: 10.1111/joor.13383
• Languages: English
• ISSNs: 0305-182X

Abstract

Abstract: Background: Temporomandibular disorders (TMDs) are disabling conditions with a negative impact on the quality of life. Their diagnosis is a complex and multi‐factorial process that should be conducted by experienced professionals, and most TMDs remain often undetected. Increasing the awareness of un‐experienced dentists and supporting the early TMD recognition may help reduce this gap. Artificial intelligence (AI) allowing both to process natural language and to manage large knowledge bases could support the diagnostic process. Objective: In this work, we present the experience of an AI‐based system for supporting non‐expert dentists in early TMD recognition. Methods: The system was based on commercially available AI services. The prototype development involved a preliminary domain analysis and relevant literature identification, the implementation of the core cognitive computing services, the web interface and preliminary testing. Performance evaluation included a retrospective review of seven available clinical cases, together with the involvement of expert professionals for usability testing. Results: The system comprises one module providing possible diagnoses according to a list of symptoms, and a second one represented by a question and answer tool, based on natural language. We found that, even when using commercial services, the training guided by experts is a key factor and that, despite the generally positive feedback, the application's best target is untrained professionals. Conclusion: We provided a preliminary proof of concept of the feasibility of implementing an AI‐based system aimed to support non‐specialists in the early identification of TMDs, possibly allowing a faster and more frequent referral to second‐level medical centres. Our results showed that AI is a useful tool to improve TMD detection by facilitating a primary diagnosis. Abstract : In this study, we present the experience of Artificial Intelligence (AI)‐based system for supporting non‐expert dentists in early Temporomandibular Disorders (TMDs) recognition based on the use of AI natural language and management of large knowledge. We presented a preliminary proof of concept "the feasibility of implementing an AI‐based system aimed to support non‐specialists in the early identification of TMDs".