Towards Trustworthy AI in Dentistry.

Area of Science:

• Medical Artificial Intelligence
• Dental Artificial Intelligence
• AI Trustworthiness

Background:

• Trust is essential for the successful implementation, acceptance, and maintenance of artificial intelligence (AI) in healthcare.
• Standardization efforts aim to improve AI quality and reliability across various attributes.
• Explainable AI (XAI) methods can enhance the transparency and trustworthiness of AI models.

Purpose of the Study:

• To review ongoing research and standardization activities contributing to trustworthy medical and dental AI.
• To discuss the role and key elements of standardization in AI development.
• To explore how explainable AI can foster trust in dental AI applications.

Main Methods:

• Literature review of current research and standardization initiatives in medical and dental AI.
• Analysis of standardization strategies for AI quality and reliability.
• Demonstration of explainable AI for a specific dental use case (caries prediction).

Main Results:

• Standardization is a key strategy for building trust in AI by ensuring quality and reliability.
• Explainable AI methods offer practical benefits for developing trustworthy AI models in dentistry.
• The use of XAI in caries prediction using near-infrared light transillumination images illustrates practical advantages.

Conclusions:

• Implementing standardization and explainable AI is vital for advancing the adoption of trustworthy AI in dentistry.
• Explainable AI provides practical solutions for enhancing the reliability and user acceptance of dental AI tools.
• Continued focus on standardization and explainable AI will drive the responsible innovation of AI in healthcare.