Non-Cochrane · DTA · PM41914450
Artificial Intelligence in Dental Treatment Planning and Diagnostic Decision-Making: A Systematic Review and Meta-Analysis.
Alabdulkareem Mohammad, Atieh Momen, AbuMostafa Ammar, et al. · Clinical and experimental dental research · 2026 · PubMed · doi:10.1002/cre2.70343
Findings
11 findings
| Population | ||||
|---|---|---|---|---|
| Dental images (radiographs, CBCT, intraoral photographs) of patients of any age | AI algorithms (deep learning, mainly CNN/YOLO) for dental detection tasks | Reference standard (expert clinicians or consensus) | Pooled F1 score Dental imaging (radiographs, CBCT, intraoral photographs) | F1 score 0.90 [0.77, 0.96] Moderate Downgraded: inconsistency 8 studies · 24033 participants |
| Dental images (radiographs, CBCT, intraoral photographs) of patients of any age | AI algorithms (deep learning, mainly CNN/YOLO) for dental detection tasks | Reference standard (expert clinicians or consensus) | Pooled precision Dental imaging (radiographs, CBCT, intraoral photographs) | Precision 0.88 [0.71, 0.96] Moderate Downgraded: risk of bias, inconsistency 7 studies · 23248 participants |
| Patients undergoing dental imaging or diagnostic assessment | AI algorithms for dental detection tasks | Reference standard (expert clinicians) | Pooled specificity Dental imaging (panoramic, periapical, bitewing and CBCT radiographs, intraoral photographs) | Specificity (proportion) 0.94 [0.86, 0.97] Moderate Downgraded: risk of bias, inconsistency 10 studies · 21516 participants |
| Patients undergoing dental imaging or diagnostic assessment | AI algorithms for dental detection tasks | Reference standard (expert clinicians) | Pooled area under the curve Dental imaging (panoramic, periapical, bitewing and CBCT radiographs, intraoral photographs) | AUC 0.93 GRADE not reported Studies not reported |
| Patients undergoing dental imaging or diagnostic assessment | AI algorithms for dental detection tasks | Reference standard (expert clinicians) | Pooled sensitivity Dental imaging (panoramic, periapical, bitewing and CBCT radiographs, intraoral photographs) | Sensitivity (proportion) 0.85 [0.76, 0.91] Moderate Downgraded: risk of bias, inconsistency 8 studies · 11452 participants |
| Patients undergoing dental imaging or diagnostic assessment | AI algorithms for dental segmentation tasks | Expert annotation (reference standard) | Precision in segmentation tasks (risk difference) Dental imaging (panoramic, periapical, bitewing and CBCT radiographs, intraoral photographs) | RD 0.86 [-0.73, 2.46] GRADE not reported 2 studies |
| Dental images (radiographs, CBCT) in segmentation tasks | AI segmentation algorithms (e.g. U-Net, YOLO, SAM) | Expert annotation (reference standard) | Risk difference for precision in segmentation tasks Dental imaging (radiographs, CBCT, intraoral photographs) · Common effect model | RD 0.98 [0.97, 0.98] GRADE not reported 2 studies · 4039 participants |
| Dental images (radiographs, CBCT) in segmentation tasks | AI segmentation algorithms (e.g. U-Net, YOLO, SAM) | Expert annotation (reference standard) | Pooled Dice similarity coefficient (DSC/F1) Dental imaging (radiographs, CBCT, intraoral photographs) | Dice similarity coefficient 0.89 [0.13, 1.00] Low Downgraded: imprecision, inconsistency 3 studies · 4539 participants |
| Clinicians interpreting dental radiographs | AI assistance in clinical decision-making | Usual clinician decision-making without AI | Interobserver agreement with AI assistance Dental imaging (panoramic, periapical, bitewing and CBCT radiographs, intraoral photographs) | Not applicable (effect sizes could not be calculated); 4 studies GRADE not reported 4 studies |
| Clinicians interpreting dental radiographs | AI assistance in clinical decision-making | Usual clinician decision-making without AI | Diagnostic efficiency (time reduction) Dental imaging (panoramic, periapical, bitewing and CBCT radiographs, intraoral photographs) | Not applicable (effect sizes could not be calculated); 2 studies Low Downgraded: imprecision 2 studies |
| Patients undergoing dental treatment planning | AI treatment planning support | Usual treatment planning without AI | Treatment planning support Dental imaging (panoramic, periapical, bitewing and CBCT radiographs, intraoral photographs) | Not applicable (effect sizes could not be calculated); 3 studies Low Downgraded: indirectness 3 studies |
Included primary studies 27
Open in Studies| Study | Year | Also included by | Outcomes it informs here | Trial data |
|---|---|---|---|---|
| Vinayahalingam 2021 Classification of caries in third molars on panoramic radiographs using deep learning. PMID:34131266 | 2021 | 2 other reviews | Pooled sensitivity; Pooled specificity | not extracted yet |
| Kurt 2024 Evaluation of tooth development stages with deep learning-based artificial intelligence algorithm. PMID:39227802 | 2024 | 1 other review | Pooled F1 score; Pooled precision | not extracted yet |
| Zhang 2024 Diagnostic accuracy of artificial intelligence-assisted caries detection: a clinical evaluation. PMID:39285427 | 2024 | 1 other review | Pooled sensitivity; Pooled specificity; Pooled F1 score; Pooled precision | not extracted yet |
| Alotaibi 2022 Artificial intelligence (AI) diagnostic tools: utilizing a convolutional neural network (CNN) to assess periodontal bone level… PMID:36100856 | 2022 | this review only | Pooled sensitivity; Pooled specificity | not extracted yet |
| Bayati 2025 Advanced AI-driven detection of interproximal caries in bitewing radiographs using YOLOv8. PMID:39920198 | 2025 | this review only | Pooled F1 score; Pooled precision | not extracted yet |
| Bayrakdar 2021 A deep learning approach for dental implant planning in cone-beam computed tomography images. PMID:34011314 | 2021 | this review only | not stated | not extracted yet |
| Beser 2024 Correction: YOLO-V5 based deep learning approach for tooth detection and segmentation on pediatric panoramic radiographs in mixed dentition. PMID:39198729 | 2024 | this review only | Pooled F1 score; Pooled precision; Pooled Dice similarity coefficient (DSC/F1); Risk difference for precision in segmentation tasks | not extracted yet |
| Celik 2022 Automated detection of dental restorations using deep learning on panoramic radiographs. PMID:36043433 | 2022 | this review only | not stated | not extracted yet |
| Chau 2025 A novel AI model for detecting periapical lesion on CBCT: CBCT-SAM. PMID:39667487 | 2025 | this review only | Pooled sensitivity; Pooled specificity; Pooled Dice similarity coefficient (DSC/F1); Risk difference for precision in segmentation tasks | not extracted yet |
| Chen 2023 Automatic recognition of teeth and periodontal bone loss measurement in digital radiographs using deep-learning artificial intelligence. PMID:37404656 | 2023 | this review only | not stated | not extracted yet |
| Esmaeilyfard 2023 Dental Caries Detection and Classification in CBCT Images Using Deep Learning. PMID:37940474 | 2024 | this review only | Pooled specificity; Pooled F1 score | not extracted yet |
| Frutos 2024 AI-Dentify: deep learning for proximal caries detection on bitewing x-ray - HUNT4 Oral Health Study. PMID:38494481 | 2024 | this review only | not stated | not extracted yet |
| Ibraheem 2025 Assessment of the Diagnostic Accuracy of Artificial Intelligence Software in Identifying Common Periodontal and Restorative Dental… PMID:40507004 | 2025 | this review only | Pooled sensitivity; Pooled specificity | not extracted yet |
| Kazimierczak 2024 Periapical Lesions in Panoramic Radiography and CBCT Imaging-Assessment of AI's Diagnostic Accuracy. PMID:38731237 | 2024 | this review only | not stated | not extracted yet |
| Leemput 2024 matched by author and year (probable) | NR | this review only | Pooled sensitivity; Pooled specificity | not extracted yet |
| Li 2022 Artificial intelligence for caries and periapical periodontitis detection. PMID:35341892 | 2022 | this review only | Pooled sensitivity; Pooled specificity; Pooled F1 score; Pooled precision | not extracted yet |
| Mun 2024 Deep learning-based prediction of indication for cracked tooth extraction using panoramic radiography. PMID:39152384 | 2024 | this review only | not stated | not extracted yet |
| Muresanu 2024 Automating Dental Condition Detection on Panoramic Radiographs: Challenges, Pitfalls, and Opportunities. PMID:39451659 | 2024 | this review only | not stated | not extracted yet |
| Parenti 2025 Artificial Intelligence-Aided Tooth Detection and Segmentation on Pediatric Panoramic Radiographs in Mixed Dentition Using a Transfer… PMID:41153287 | 2025 | this review only | not stated | not extracted yet |
| Peker 2025 Evaluation of the Performance of a YOLOv10-Based Deep Learning Model for Tooth Detection and Numbering on Panoramic Radiographs of… PMID:40002557 | 2025 | this review only | Pooled F1 score; Pooled precision | not extracted yet |
| Perez 2023 matched by author and year (probable) | NR | this review only | not stated | not extracted yet |
| Pornprasertsuk-Damrongsri 2025 Clinical application of deep learning for enhanced multistage caries detection in panoramic radiographs. PMID:41022932 | 2025 | this review only | Pooled specificity; Pooled Dice similarity coefficient (DSC/F1) | not extracted yet |
| Saber 2025 Automated assessment of periapical health based on the radiographic periapical index using YOLOv8, YOLOv11, and YOLOv12 one-stage object… PMID:41115912 | 2025 | this review only | Pooled F1 score; Pooled precision | not extracted yet |
| Wang 2025 DOI:10.48550/arxiv.2502.10277 | NR | this review only | not stated | not extracted yet |
| Xue 2024 Deep learning method to automatically diagnose periodontal bone loss and periodontitis stage in dental panoramic radiograph. PMID:39332519 | 2024 | this review only | not stated | not extracted yet |
| Yoon 2024 AI-based dental caries and tooth number detection in intraoral photos: Model development and performance evaluation. PMID:38145804 | 2024 | this review only | not stated | not extracted yet |
| Zhu 2023 Artificial intelligence in the diagnosis of dental diseases on panoramic radiographs: a preliminary study. PMID:37270488 | 2023 | this review only | Pooled sensitivity; Pooled specificity | not extracted yet |
Overlap with other reviews
Reviews that include the same primary studies are not independent evidence. Shared counts match studies by PMID, DOI or NCT, and otherwise by author and year (a probable match). Only the non-Cochrane dental reviews have study lists, so overlap with Cochrane reviews is not counted yet.
Review question
| Element | Planned (methods) | Found (results) |
|---|---|---|
| Population | Patients of any age undergoing dental imaging, diagnostic assessment or treatment planning across all dental specialties | 60,857 radiographic images from 27 studies (panoramic most common, then periapical, bitewing, CBCT, intraoral photographs); ages 5 to 95 years; caries, periodontal conditions, periapical lesions, tooth detection and numbering, restorations, impacted and cracked teeth |
| Intervention | AI algorithms (machine learning, deep learning, classical ML) used to assist or perform diagnostic tasks or support treatment planning | Deep CNNs, mostly YOLO variants (YOLOv5, v8, v10, v11), U-Net, transfer learning models (VGG, ResNet, MobileNet, InceptionV3) (n=27) |
| Comparison | Reference standard (expert clinicians, histopathology, clinical follow-up) or human clinicians; usual clinician decision-making without AI or with alternative aids | Reference standard of expert clinician or consensus annotation; unassisted clinicians in some studies |
| Outcomes | Sensitivity, specificity, AUC, precision, recall, F1, accuracy, DOR; decision-making outcomes (treatment plan changes, inter-rater reliability, time to decision, clinician confidence, patient-level outcomes) | Pooled sensitivity, specificity, F1, precision (detection; 15 studies in meta-analysis), Dice similarity coefficient (segmentation), precision risk difference; narrative interobserver agreement, time reduction, treatment planning support |
| Study designs | Diagnostic accuracy studies, validation studies, RCTs of AI vs usual care, interventional and observational studies of clinician-AI interaction, algorithm development papers with independent validation | 27 studies, mostly retrospective diagnostic accuracy studies; 15 in the quantitative meta-analysis, 12 narrative only |
| Timing | Not specified | Not specified |