Can Dental AI Really Beat Dentists? DentalPair-Cert for Rigorous AI-Dentist Inference
- 1The University of Texas at Dallas
Abstract
A dental AI system and a dentist reading the same radiographs form a paired comparison. Published comparative studies often report the two arms separately against a reference standard, leaving the joint pattern of correctness between them unavailable for secondary paired inference. We show what that omission costs. The accuracy difference remains exactly identified; its sampling variance does not, so the report contains the estimate and not its uncertainty. On a study of 282 units, two published accuracies are consistent with 38 distinct joint tables whose confidence intervals differ in width by a factor of 2.5. The consequence is a three-zone decision map rather than a single threshold: differences at or below 1.06 points are non-significant under every compatible table, differences at or above 6.03 points are significant under every compatible table, and in between the published numbers cannot decide. We then show the omission is repairable at negligible cost. One additional integer, the number of units both arms classify correctly, identifies the joint table exactly and restores standard paired inference. For a panel of readers the pairwise dependences must arise from one joint distribution, a constraint that binds once three readers are present; publishing each readers joint-correct count against a single reference reader cannot widen and may tighten every pairwise bound, and in a 7-arm experiment reduced them by a median of 37% even for pairs excluding that reference. Where the integer was never published we give DentalPair-Cert, an interval with finite-sample coverage uniformly over every admissible within-unit AI-dentist dependence under the independent-sampling-unit model, certified in both the nuisance maximisation and the inversion. Across 4,200,000 simulated comparisons an independence analysis falls to 74.5% coverage with 12.2% type-I error; in a purposive sample of 9 recent comparative studies, 1 reported a paired test on discordant units.
Related
- Operators matter \textendash An assessment of the expectations, perceptions, and performance of dentists, postgraduate students, and dental prosthetist students using intraoral scanning
- Dentists' and Dental Assistants' Perceptions of the New Profession of Dental Hygienist in France
- Knowledge and awareness of sensory processing disorder and sensory-adapted dental care among pediatric dentists: A cross-sectional survey
- Involving Independent Mental Capacity Advocates (IMCAs) in Dental Care Decision-Making: A Qualitative Exploration of Dentists' Experiences in England and Wales.
- Case-Based Diagnostic Performance, Challenges and Tool Development in Identifying Orofacial Conditions Among Pediatric Dentists: A Mixed-Methods Study.
BibTeX
@misc{alve2026dental,
title = {{Can Dental AI Really Beat Dentists? DentalPair-Cert for Rigorous AI-Dentist Inference}},
author = {S. R. Alve and S. Rahman and S. M. A. C. Meem},
howpublished = {medRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.01.26361874},
url = {https://www.medrxiv.org/content/10.64898/2026.09.01.26361874v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Can Dental AI Really Beat Dentists? DentalPair-Cert for Rigorous AI-Dentist Inference
AU - S. R. Alve
AU - S. Rahman
AU - S. M. A. C. Meem
PY - 2026
DA - 2026/09/01
PB - medRxiv
DO - 10.64898/2026.09.01.26361874
UR - https://www.medrxiv.org/content/10.64898/2026.09.01.26361874v1
AB - A dental AI system and a dentist reading the same radiographs form a paired comparison. Published comparative studies often report the two arms separately against a reference standard, leaving the joint pattern of correctness between them unavailable for secondary paired inference. We show what that omission costs. The accuracy difference remains exactly identified; its sampling variance does not, so the report contains the estimate and not its uncertainty. On a study of 282 units, two published accuracies are consistent with 38 distinct joint tables whose confidence intervals differ in width by a factor of 2.5. The consequence is a three-zone decision map rather than a single threshold: differences at or below 1.06 points are non-significant under every compatible table, differences at or above 6.03 points are significant under every compatible table, and in between the published numbers cannot decide. We then show the omission is repairable at negligible cost. One additional integer, the number of units both arms classify correctly, identifies the joint table exactly and restores standard paired inference. For a panel of readers the pairwise dependences must arise from one joint distribution, a constraint that binds once three readers are present; publishing each readers joint-correct count against a single reference reader cannot widen and may tighten every pairwise bound, and in a 7-arm experiment reduced them by a median of 37% even for pairs excluding that reference. Where the integer was never published we give DentalPair-Cert, an interval with finite-sample coverage uniformly over every admissible within-unit AI-dentist dependence under the independent-sampling-unit model, certified in both the nuisance maximisation and the inversion. Across 4,200,000 simulated comparisons an independence analysis falls to 74.5% coverage with 12.2% type-I error; in a purposive sample of 9 recent comparative studies, 1 reported a paired test on discordant units.
ER -