Non-Cochrane · DTA · PM42645587
Detection of oral potentially malignant disorders and oral cancer with white light through deep learning: a systematic review and meta-analysis.
Soto Ayén N, Rodríguez Molinero J A, Navarro Lorente P J, et al. · Clinical oral investigations · 2026 · PubMed · doi:10.1007/s00784-026-07058-5
Findings
14 findings
| Population | ||||
|---|---|---|---|---|
| Patients of any age with oral potentially malignant disorders (OPMD); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Sensitivity | Sensitivity 0.81 [0.728, 0.872] GRADE not reported 12 studies |
| Patients of any age with oral potentially malignant disorders (OPMD); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Specificity | Specificity (univariate random-effects proportion) 0.84 [0.77, 0.89] GRADE not reported Studies not reported |
| Patients of any age with oral potentially malignant disorders (OPMD); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Sensitivity | Sensitivity (univariate random-effects proportion) 0.81 [0.71, 0.88] GRADE not reported Studies not reported |
| Patients of any age with oral potentially malignant disorders (OPMD); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Specificity | Specificity (univariate common-effect proportion) 0.79 [0.79, 0.80] GRADE not reported Studies not reported |
| Patients of any age with oral potentially malignant disorders (OPMD); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Sensitivity | Sensitivity (univariate common-effect proportion) 0.76 [0.75, 0.76] GRADE not reported Studies not reported |
| Patients of any age with oral potentially malignant disorders (OPMD); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Specificity | Specificity 0.161 [0.113, 0.223] GRADE not reported 12 studies |
| Patients of any age with oral cancer (OSCC); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Sensitivity | Sensitivity (univariate random-effects proportion) 0.84 [0.72, 0.91] GRADE not reported Studies not reported |
| Patients of any age with oral cancer (OSCC); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Specificity | Specificity (univariate common-effect proportion) 0.82 [0.81, 0.82] GRADE not reported Studies not reported |
| Patients of any age with oral cancer (OSCC); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Sensitivity | Sensitivity 0.836 [0.730, 0.906] GRADE not reported 12 studies |
| Patients of any age with oral cancer (OSCC); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Specificity | Specificity (univariate random-effects proportion) 0.86 [0.77, 0.92] GRADE not reported Studies not reported |
| Patients of any age with oral cancer (OSCC); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Specificity | Specificity 0.138 [0.081, 0.226] GRADE not reported 12 studies |
| Patients of any age with oral cancer (OSCC); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Diagnostic odds ratio | DOR 0.814 [0.329, 2.012] GRADE not reported 12 studies |
| Patients of any age with oral cancer (OSCC); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Area under the SROC curve | AUC 0.994 GRADE not reported 12 studies |
| Patients of any age with oral cancer (OSCC); white-light clinical photographs | Deep learning algorithms applied to white-light digital photographs (index test) | Reference standard (histology or expert clinical assessment); no comparator defined | Sensitivity | Sensitivity (univariate common-effect proportion) 0.78 [0.77, 0.79] GRADE not reported Studies not reported |
Included primary studies 22
Open in Studies| Study | Year | Also included by | Outcomes it informs here | Trial data |
|---|---|---|---|---|
| Fu 2020 A deep learning algorithm for detection of oral cavity squamous cell carcinoma from photographic images: A retrospective study DOI:10.1016/j.eclinm.2020.100558 | 2020 | 1 other review | not stated | not extracted yet |
| Jubair 2022 A novel lightweight deep convolutional neural network for early detection of oral cancer DOI:10.1111/odi.13825 | 2021 | 1 other review | not stated | not extracted yet |
| Warin 2022 (classification and detection of OPMD) Performance of deep convolutional neural network for classification and detection of oral potentially malignant disorders in photographic… DOI:10.1016/j.ijom.2021.09.001 | 2022 | 1 other review | not stated | not extracted yet |
| Dissanayake 2024 Multimodal Deep Convolutional Neural Network Pipeline for AI-Assisted Early Detection of Oral Cancer DOI:10.1109/access.2024.3454338 | 2024 | this review only | not stated | not extracted yet |
| Ferrer-Sanchez 2022 Prediction of the risk of cancer and the grade of dysplasia in leukoplakia lesions using deep learning DOI:10.1016/j.oraloncology.2022.105967 | 2022 | this review only | not stated | not extracted yet |
| Figueroa 2022 Interpretable deep learning approach for oral cancer classification using guided attention inference network DOI:10.1117/1.jbo.27.1.015001 | 2022 | this review only | not stated | not extracted yet |
| Keser 2022 A deep learning algorithm for classification of oral lichen planus lesions from photographic images: A retrospective study DOI:10.1016/j.jormas.2022.08.007 | 2023 | this review only | not stated | not extracted yet |
| Kouketsu 2024 Detection of oral cancer and oral potentially malignant disorders using artificial intelligence‐based image analysis DOI:10.1002/hed.27843 | 2024 | this review only | not stated | not extracted yet |
| Maghsoudi 2013 Disseminated Penicillium Marneffei infection with verrucoid lesions in an AIDS patient in Beijing, a non-endemic region. PMID:20133218 | 2010 | this review only | not stated | not extracted yet |
| Maghsoudi 2013 Horses Failed to Learn from Humans by Observation. PMID:32013218 | 2020 | this review only | not stated | not extracted yet |
| Rabinovici-Cohen 2024 From Pixels to Diagnosis: Algorithmic Analysis of Clinical Oral Photos for Early Detection of Oral Squamous Cell Carcinoma. PMID:38473377 | 2024 | this review only | not stated | not extracted yet |
| Saldivia-Siracusa 2025 Automated classification of oral potentially malignant disorders and oral squamous cell carcinoma using a convolutional neural network… DOI:10.1016/j.lana.2025.101138 | 2025 | this review only | not stated | not extracted yet |
| Schwarzler 2025 Machine learning versus clinicians for detection and classification of oral mucosal lesions DOI:10.1016/j.jdent.2025.105992 | 2025 | this review only | not stated | not extracted yet |
| Shawky 2024 Early Diagnosis of Oral Cancer Using Image Processing and Artificial Intelligence DOI:10.54216/fpa.140122 | 2024 | this review only | not stated | not extracted yet |
| Song 2021 (imbalanced data) Classification of imbalanced oral cancer image data from high-risk population DOI:10.1117/1.jbo.26.10.105001 | 2021 | this review only | not stated | not extracted yet |
| Song 2021 (VGG-19) Bayesian deep learning for reliable oral cancer image classification DOI:10.1364/boe.432365 | 2021 | this review only | not stated | not extracted yet |
| Song 2022 (U-Net) Exploring uncertainty measures in convolutional neural network for semantic segmentation of oral cancer images DOI:10.1117/1.jbo.27.11.115001 | 2022 | this review only | not stated | not extracted yet |
| Talwar 2023 AI-Assisted Screening of Oral Potentially Malignant Disorders Using Smartphone-Based Photographic Images. PMID:37627148 | 2023 | this review only | not stated | not extracted yet |
| Vinayahalingam 2024 Advancements in diagnosing oral potentially malignant disorders: leveraging Vision transformers for multi-class detection DOI:10.1007/s00784-024-05762-8 | 2024 | this review only | not stated | not extracted yet |
| Warin 2022 (novel deep CNNs) AI-based analysis of oral lesions using novel deep convolutional neural networks for early detection of oral cancer DOI:10.1371/journal.pone.0273508 | 2022 | this review only | not stated | not extracted yet |
| Wuttisarnwattana 2024 Precise Identification of Oral Cancer Lesions Using Artificial Intelligence DOI:10.3233/shti240601 | 2024 | this review only | not stated | not extracted yet |
| Ye 2024 Utilizing deep learning for automated detection of oral lesions: A multicenter study DOI:10.1016/j.oraloncology.2024.106873 | 2024 | this review only | not stated | 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 | Shared studies | Jaccard |
|---|---|---|
| Artificial intelligence for oral cancer diagnosis: a systematic review and meta-analysis of image-based and non-imaging models. (2026) | 3 | 0.06 |
Review question
| Element | Planned (methods) | Found (results) |
|---|---|---|
| Population | Patients of any age with a diagnosis of OPMD or oral cancer | Patients with OPMD or oral cancer photographed under white light (21 studies; 12 in meta-analysis) |
| Intervention | Deep learning to diagnose OPMD or oral cancer from digital photographic images | Deep learning models (ResNet, VGG19, DenseNet, YOLO, U-Net, others); smartphone or digital camera images |
| Comparison | No control group defined | Histological control in 11 of 21 studies; otherwise expert clinical assessment or annotation |
| Outcomes | Accuracy of deep learning (sensitivity, specificity, TP, FP, FN, TN, PPV, NPV, F1, AUC) | Pooled sensitivity and specificity (bivariate Reitsma), DOR and SROC AUC for oral cancer (n=12 studies) |
| Study designs | Original articles, clinical trials, observational studies; prospective or retrospective | Retrospective and cross-sectional diagnostic accuracy studies (n=21) |
| Timing | Not specified | Not specified |