Related Experiment Video
Updated: Jun 19, 2026
Reliability of Artificial Intelligence-Based Cone Beam Computed Tomography Integration with Digital Dental Images
Published on: February 23, 2024
Artificial intelligence applications in automated dental report generation - a scoping review
Madeline Yon1, Ethan Ng1,2, Michael M Bornstein3
1Centre for Oral Clinical Research, Institute of Dentistry, Faculty of Medicine and Dentistry, Queen Mary University of London (QMUL), London, United Kingdom.
Background:
Artificial intelligence (AI), particularly large language models and natural language processing (NLP), has enabled the structured synthesis and summarisation of complex clinical data. In healthcare, AI-driven report generation has the potential to improve efficiency, reduce clinician workload, and enhance communication with patients and healthcare professionals. This scoping review aimed to map current developments in AI applications for dental report generation and to identify existing research gaps.
Materials And Methods:
A systematic search of Embase, MEDLINE via Ovid, and IEEE Xplore was conducted for studies published between January 2015 and March 2026. Eligible studies described development, evaluation, or application of AI systems for generating dental reports from imaging, textual, or voice inputs. Extracted data on AI models, input modality, language, and evaluation methods were summarised descriptively.
Results:
1,265 records were identified, of which seven studies met inclusion criteria. Six focused on radiology report generation from panoramic radiographs, and one generated clinical examination reports from voice-transcribed charting. NLP-based models, including GPT variants and fine-tuned large language models, were used to generate final reports. Performance was evaluated using metrics including ROUGE, BLEU, BERTScore, hallucination analysis, structural validity, response latency, output length, readability indices, clinician ratings, and patient questionnaires. AI-generated reports demonstrated high accuracy for common findings and readability comparable to human-authored reports. Simplified AI-adapted versions improved patient-rated clarity. However, heterogeneity in report types, datasets, languages, and evaluation metrics limited direct comparison.
Conclusion:
AI systems show promising capability in generating dental reports from diverse inputs and customisation for professional or non-professional audience. Standardised evaluation frameworks, larger multilingual datasets, and assessment of patient comprehension are needed before routine clinical implementation.
