Recent Highlights in Dental AI Research (March 1, 2026 – July 1, 2026)
Every four months, the AAAI-D Council for Research and Evidence reviews recently published literature and selects studies that may be of particular interest to clinicians, educators, researchers, and others following developments in dental artificial intelligence. Selection is based on factors such as study design, methodological approach, clinical relevance, or the questions a study raises for future investigation. These summaries are intended to encourage further reading rather than replace the original publications.
1. Demographic Prompt Cues Shift Clinical Recommendations in Multimodal LLMs
Babak Saravi, et al. BMC Oral Health. 2026. DOI: 10.1186/s12903-026-08644-5
What they did: A multi-model audit of 25,380 dental radiograph assessments examined whether demographic cues embedded within prompts (age, sex, ethnicity) altered clinical recommendations generated by multimodal large language models using a public radiographic dataset.
Why it matters: Studies examining potential sources of bias remain relatively limited. This work highlights that consistency between AI systems deserves attention alongside diagnostic performance.
2. LLMs vs. Prosthodontic Residents on Board Examinations
Soni Prasad, et al. The Journal of Prosthetic Dentistry. 2026. DOI: 10.1016/j.prosdent.2026.03.043
What they did: Three large language models were compared with prosthodontic residents using the 2024 and 2025 National Prosthodontic Resident Examinations (300 multiple-choice questions).
Why it matters: Comparisons using specialty board examinations remain relatively uncommon in the dental AI literature. The study provides a practical benchmark that many educators and clinicians can readily interpret.
3. ML Model Predicts Peri-Implant Mucositis at One Year
Lin Liu, et al. Clinical Oral Implants Research. 2026. DOI: 10.1111/clr.70125
What they did: A support vector machine model built on ten clinical variables was externally validated to predict peri-implant mucositis one year after implant placement.
Why it matters: External validation remains less common than internal testing in many AI studies. Reports that examine performance beyond the original development dataset provide useful perspective on generalizability.
4. Deep Learning Detects Endodontic-Origin Maxillary Sinusitis on CBCT
Omar Ayman Saleh Sherif, et al. Scientific Reports. 2026. DOI: 10.1038/s41598-026-52147-w
What they did: A multi-stage deep learning pipeline classified maxillary sinus status on CBCT scans and was evaluated using both internal and independent external datasets.
Why it matters: The inclusion of an independent external dataset distinguishes this study from many earlier imaging reports and contributes to the growing body of externally validated research.
5. AI Teleorthodontic Triage Compared with Conventional Assessment
Maxim Milosevic, et al. European Journal of Orthodontics. 2026. DOI: 10.1093/ejo/cjag050
What they did: A crossover randomized clinical trial compared AI-assisted teleorthodontic triage with conventional face-to-face referral assessment within a publicly funded healthcare system.
Why it matters: Prospective randomized studies remain comparatively uncommon in dental AI and offer a different perspective than retrospective algorithm evaluations.
6. ML Triage Score for Medication-Related Osteonecrosis of the Jaw (MRONJ)
Hui One Jeong, et al. Diagnostics. 2026. DOI: 10.3390/diagnostics16121887
What they did: A six-variable machine learning-derived triage score for MRONJ risk in osteoporosis patients undergoing extraction was evaluated in an independent multicenter cohort.
Why it matters: The study reflects continued interest in applying AI to clinical decision support beyond image interpretation alone.
7. AI-Personalized Video Reduces Pediatric Dental Anxiety — With Caveats
A. Tasgaonkar, et al. European Archives of Paediatric Dentistry. 2026. DOI: 10.1007/s40368-026-01241-8
What they did: A randomized controlled trial compared AI-personalized video self-modeling with conventional audiovisual preparation for reducing dental anxiety in children.
Why it matters: Studies reporting both positive and neutral findings help define where AI personalization may provide measurable benefit and where additional investigation is still warranted.
8. Q-Bone: Multicenter-Validated System for Alveolar Bone Loss Quantification
Wei Li, et al. Journal of Translational Medicine. 2026. DOI: 10.1186/s12967-026-08444-z
What they did: An AI system for automated tooth segmentation, landmark localization, and alveolar bone loss quantification was validated across four centers and multiple imaging platforms against specialist reference standards.
Why it matters: Multicenter validation provides a broader perspective on how imaging algorithms may perform across different clinical settings and imaging environments.
9. Systematic Review: AI in Orthodontic Treatment Planning
Martin Baxmann, et al. BMC Oral Health. 2026. DOI: 10.1186/s12903-026-08854-x
What they did: A registered systematic review synthesized published evidence evaluating AI and knowledge-based systems for orthodontic treatment planning, incorporating formal risk-of-bias assessment.
Why it matters: Systematic reviews that evaluate study quality alongside reported performance help place individual findings within the broader evidence base.
10. AI Crown Design: Faster, But Not Better
Luiz Felipe Fernandes Gonçalves, et al. Journal of Dentistry. 2026. DOI: 10.1016/j.jdent.2026.106755
What they did: An in vitro comparison evaluated AI-generated single-unit crown designs against conventional CAD workflows for design time, marginal fit, morphology, and occlusal contacts.
Why it matters: Efficiency gains are important, but they should be considered separately from restoration quality. Studies that report both outcomes provide a more complete assessment of AI-assisted workflows.
Prepared by the AAAI-D Council for Research and Evidence
The Council periodically reviews recently published literature to identify studies that contribute to the evolving evidence base in dental artificial intelligence. Inclusion in this summary does not constitute endorsement by the American Academy of Artificial Intelligence in Dentistry (AAAI-D).