Applications and progress of AI in mechanistic research and clinical diagnosis of oral genetic and rare diseases

doi:10.3969/j.issn.1674-8115.2026.07.001

Abstract

Abstract:

Oral genetic and rare diseases are characterized by low prevalence, marked phenotypic heterogeneity, and complex pathogenic mechanisms. They often present with severe clinical manifestations, pose significant diagnostic challenges, and lack effective therapeutic options. With the rapid advancement of multi‑omics sequencing and medical imaging technologies, large‑scale molecular and clinical data have provided new perspectives for understanding these disorders. Artificial intelligence (AI), with its powerful capabilities in pattern recognition and modeling complex relationships, has achieved significant progress in both fundamental research and clinical diagnosis and treatment related to oral genetic and rare diseases. In studies of pathogenic mechanisms, AI enables the explanation of genomic, transcriptomic, proteomic, and microbiome data, facilitates the identification of novel molecular biomarkers, and supports the construction of disease prediction models, thereby advancing the elucidation of complex etiologies. In clinical diagnosis, AI significantly enhances the automation and accuracy of oral imaging analysis, enabling intelligent diagnosis of dental caries, tooth developmental anomalies, salivary gland diseases, fibrous dysplasia, and various craniofacial malformations. AI‑based tools also provide decision support for surgical treatment planning. Despite these advances, challenges remain, including data scarcity, limited model interpretability, and ethical and regulatory concerns. This paper aims to provide a comprehensive review of the latest advancements in the application of AI in the field of oral genetic and rare diseases, systematically examining the progress, challenges, and prospects in deepening the understanding of disease mechanisms and accelerating clinical diagnosis, with the goal of ultimately promoting the overall advancement of precision medicine for oral genetic diseases and improving the accessibility of high-quality healthcare resources.

Key words: artificial intelligence (AI), oral genetic and rare disease, pathogenic mechanism, clinical diagnosis, precision medicine

CLC Number:

R782

Shen Jingting, Liu Zhixu, Wang Xudong. Applications and progress of AI in mechanistic research and clinical diagnosis of oral genetic and rare diseases[J]. Journal of Shanghai Jiao Tong University (Medical Science), 2026, 46(7): 829-838.

Figures/Tables 3

TrendMD

References 53

[1]	Arno P S, Bonuck K, Davis M. Rare diseases, drug development, and AIDS: the impact of the Orphan Drug Act[J]. Milbank Q, 1995, 73(2): 231-252.
[2]	Franco P. Orphan drugs: the regulatory environment[J]. Drug Discov Today, 2013, 18(3/4): 163-172.
[3]	Nguengang Wakap S, Lambert D M, Olry A, et al. Estimating cumulative point prevalence of rare diseases: analysis of the Orphanet database[J]. Eur J Hum Genet, 2020, 28(2): 165-173.
[4]	Taneyhill L A, Hoover-Fong J, Lozanoff S, et al. The society for craniofacial genetics and developmental biology 38th annual meeting[J]. Am J Med Genet A, 2016, 170(7): 1732-1753.
[5]	段小红. 口腔遗传病学[M]. 北京: 人民卫生出版社, 2012.
	Duan X H. Genetic diseases of oral cavity[M]. Beijing: People's Medical Publishing House, 2012.
[6]	段小红. 口腔遗传性疾病系列讲座(四): 口腔单基因遗传病与罕见病的遗传研究策略[J]. 中华口腔医学杂志, 2015, 50(7): 442-444.
	Duan X H. Lecture series on oral hereditary diseases (Ⅳ): genetic research strategies for monogenic oral diseases and rare diseases[J]. Chinese Journal of Stomatology, 2015, 50(7): 442-444.
[7]	Beleza-Meireles A, Hart R, Clayton-Smith J, et al. Oculo-auriculo-vertebral spectrum: clinical and molecular analysis of 51 patients[J]. Eur J Med Genet, 2015, 58(9): 455-465.
[8]	Baião A R, Cai Z X, et al. A technical review of multi-omics data integration methods: from classical statistical to deep generative approaches[J]. Brief Bioinform, 2025, 26(4): bbaf355.
[9]	Banerjee J, Taroni J N, Allaway R J, et al. Machine learning in rare disease[J]. Nat Meth, 2023, 20(6): 803-814.
[10]	Chen X X, Wang X M, Zhang K, et al. Recent advances and clinical applications of deep learning in medical image analysis[J]. Med Image Anal, 2022, 79: 102444.
[11]	Zhang S J, Meng P Q, Zhang J N, et al. Machine learning models for genetic risk assessment of infants with non-syndromic orofacial cleft[J]. Genomics Proteomics Bioinformatics, 2018, 16(5): 354-364.
[12]	Maurano M T, Humbert R, Rynes E, et al. Systematic localization of common disease-associated variation in regulatory DNA[J]. Science, 2012, 337(6099): 1190-1195.
[13]	Edwards S L, Beesley J, French J D, et al. Beyond GWASs: illuminating the dark road from association to function[J]. Am J Hum Genet, 2013, 93(5): 779-797.
[14]	Trynka G, Sandor C, Han B, et al. Chromatin marks identify critical cell types for fine mapping complex trait variants[J]. Nat Genet, 2013, 45(2): 124-130.
[15]	Dai Y L, Itai T, Pei G S, et al. DeepFace: deep-learning-based framework to contextualize orofacial-cleft-related variants during human embryonic craniofacial development[J]. HGG Adv, 2024, 5(3): 100312.
[16]	Forrest I S, Vy H M T, Rocheleau G, et al. Machine learning-based penetrance of genetic variants[J]. Science, 2025, 389(6763): eadm7066.
[17]	Kamitaki N, Handsaker R E, Hujoel M L A, et al. Human and bacterial genetic variation shape oral microbiomes and health[J]. Nature, 2026, 651(8105): 429-439.
[18]	郭威, 王祥璞, 苏桐宇, 等. 单细胞测序解析先天缺牙及牙体硬组织相关分子在小鼠磨牙的时序表达特征[J]. 中华口腔医学杂志, 2025, 60(9): 987-996.
	Guo W, Wang X P, Su T Y, et al. Single-cell sequencing reveals the temporal expression characteristics of key molecules related to tooth agenesis and dental hard tissues in mouse molars[J]. Chinese Journal of Stomatology, 2025, 60(9): 987-996.
[19]	Huang W B, Qian Z W, Zhang J N, et al. A single-cell atlas of developing mouse palates reveals cellular and molecular transitions in periderm cell fate[J]. Genomics Proteomics Bioinformatics, 2025, 23(1): qzaf013.
[20]	Yankee T N, Oh S, Winchester E W, et al. Integrative analysis of transcriptome dynamics during human craniofacial development identifies candidate disease genes[J]. Nat Commun, 2023, 14(1): 4623.
[21]	Liu L J, Li W, Wang F, et al. A pre-trained large generative model for translating single-cell transcriptomes to proteomes[J]. Nat Biomed Eng, 2025: 1-20.
[22]	Thurzo A, Kurilová V, Varga I. Artificial intelligence in orthodontic smart application for treatment coaching and its impact on clinical performance of patients monitored with AI-TeleHealth system[J]. Healthcare (Basel), 2021, 9(12): 1695.
[23]	Lee J H, Kim D H, Jeong S N, et al. Detection and diagnosis of dental caries using a deep learning-based convolutional neural network algorithm[J]. J Dent, 2018, 77: 106-111.
[24]	Hiraiwa T, Ariji Y, Fukuda M, et al. A deep-learning artificial intelligence system for assessment of root morphology of the mandibular first molar on panoramic radiography[J]. Dentomaxillofac Radiol, 2019, 48(3): 20180218.
[25]	Ekert T, Krois J, Meinhold L, et al. Deep learning for the radiographic detection of apical lesions[J]. J Endod, 2019, 45(7): 917-922.e5.
[26]	Duman S, Yılmaz E F, Eşer G, et al. Detecting the presence of taurodont teeth on panoramic radiographs using a deep learning-based convolutional neural network algorithm[J]. Oral Radiol, 2023, 39(1): 207-214.
[27]	Incerti Parenti S, Tsiotas G, Maglioni A, et al. Artificial intelligence-aided tooth detection and segmentation on pediatric panoramic radiographs in mixed dentition using a transfer learning approach[J]. Diagnostics (Basel), 2025, 15(20): 2615.
[28]	Sheng C, Wang L, Huang Z H, et al. Transformer-based deep learning network for tooth segmentation on panoramic radiographs[J]. J Syst Sci Complex, 2023, 36(1): 257-272.
[29]	Felek T, Tercanlı H, Gök R Ş. Evaluating vision transformers and convolutional neural networks in the context of dental image processing: a systematic review[J]. BMC Oral Health, 2025, 25(1): 1626.
[30]	Han H, Lu F, Deng Y H, et al. M2CF-net: a multi-resolution and multi-scale cross fusion network for segmenting pathology lesion of the focal lymphocytic sialadenitis[C]//2023 IEEE International Conference on Medical Artificial Intelligence (MedAI), 2023: 425-434.
[31]	Chiesa-Estomba C M, Echaniz O, Sistiaga Suarez J A, et al. Machine learning models for predicting facial nerve palsy in parotid gland surgery for benign tumors[J]. J Surg Res, 2021, 262: 57-64.
[32]	Jurek J, Wójtowicz W, Wójtowicz A. Syntactic pattern recognition-based diagnostics of fetal palates[J]. Pattern Recognit Lett, 2020, 133: 144-150.
[33]	Shafi N, Bukhari F, Iqbal W, et al. Cleft prediction before birth using deep neural network[J]. Health Informatics J, 2020, 26(4): 2568-2585.
[34]	Kuwada C, Ariji Y, Kise Y, et al. Detection and classification of unilateral cleft alveolus with and without cleft palate on panoramic radiographs using a deep learning system[J]. Sci Rep, 2021, 11(1): 16044.
[35]	Zhang Y G, Pei Y R, Chen S, et al. Volumetric registration-based cleft volume estimation of alveolar cleft grafting procedures[C]//2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI), 2020: 99-103.
[36]	Patcas R, Bernini D A J, Volokitin A, et al. Applying artificial intelligence to assess the impact of orthognathic treatment on facial attractiveness and estimated age[J]. Int J Oral Maxillofac Surg, 2019, 48(1): 77-83.
[37]	Patcas R, Timofte R, Volokitin A, et al. Facial attractiveness of cleft patients: a direct comparison between artificial-intelligence-based scoring and conventional rater groups[J]. Eur J Orthod, 2019, 41(4): 428-433.
[38]	Huang X R, He D M, Li Z M, et al. Maxillofacial bone movements-aware dual graph convolution approach for postoperative facial appearance prediction[J]. Med Image Anal, 2025, 99: 103350.
[39]	Adel M, Hunt K J, Lau D, et al. Precision assessment of facial asymmetry using 3D imaging and artificial intelligence[J]. J Clin Med, 2025, 14(20): 7172.
[40]	Chen G J, Fu Q, Xu X, et al. Digitally guided precision fat grafting for progressive hemifacial atrophy[J]. Aesthetic Plast Surg, 2025, 49(12): 3443-3453.
[41]	Saranya A, Kottursamy K, AlZubi A A, et al. Analyzing fibrous tissue pattern in fibrous dysplasia bone images using deep R-CNN networks for segmentation[J]. Soft comput, 2022, 26(16): 7519-7533.
[42]	Li G Z, Liu H, Pan Z Y, et al. Predicting craniofacial fibrous dysplasia growth status: an exploratory study of a hybrid radiomics and deep learning model based on computed tomography images[J]. Oral Surg Oral Med Oral Pathol Oral Radiol, 2025, 139(3): 364-376.
[43]	Zhang A B, Zhao J R, Wang S, et al. CT-based radiomics analysis of different machine learning models for differentiating gnathic fibrous dysplasia and ossifying fibroma[J]. Oral Dis, 2024, 30(8): 5243-5254.
[44]	Tolkachjov S N, Patel N G, Tollefson M M. Progressive hemifacial atrophy: a review[J]. Orphanet J Rare Dis, 2015, 10: 39.
[45]	Chapurlat R D, Orcel P. Fibrous dysplasia of bone and McCune-Albright syndrome[J]. Best Pract Res Clin Rheumatol, 2008, 22(1): 55-69.
[46]	Hart E S, Kelly M H, Brillante B, et al. Onset, progression, and plateau of skeletal lesions in fibrous dysplasia and the relationship to functional outcome[J]. J Bone Miner Res, 2007, 22(9): 1468-1474.
[47]	Abdullah Al Eid R. Molecular pathogenetic options for differentiating fibrous dysplasia and ossifying fibroma: a systematically conducted literature review[J]. J Oral Maxillofac Surg Med Pathol, 2025, 37(6): 1283-1293.
[48]	Liang L G, Chen Y L, Wang T F, et al. Genetic transformer: an innovative large language model driven approach for rapid and accurate identification of causative variants in rare genetic diseases[J]. medRxiv, 2024: 2024.07.18.24310666v1. DOI: 10.1101/2024.07.18.24310666.
[49]	Song J, Xu Z C, He M Q, et al. Graph retrieval augmented large language models for facial phenotype associated rare genetic disease[J]. NPJ Digit Med, 2025, 8(1): 543.
[50]	Hao J, Fan Y X, Sun Y P, et al. Towards better dental AI: A multimodal benchmark and instruction dataset for panoramic X-ray analysis [J]. arXiv, 2025: 2509.09254v1.
[51]	Topol E J. High-performance medicine: the convergence of human and artificial intelligence[J]. Nat Med, 2019, 25(1): 44-56.
[52]	Alsentzer E, Li M M, Kobren S N, et al. Few shot learning for phenotype-driven diagnosis of patients with rare genetic diseases[J]. NPJ Digit Med, 2025, 8(1): 380.
[53]	Rieke N, Hancox J, Li W Q, et al. The future of digital health with federated learning[J]. NPJ Digit Med, 2020, 3: 119.

Reference	Year	Dataset	Objective	AI model
Zhang, et al^[11]	2018	43 GWAS-significant SNPs associated with OFCs	Genetic risk assessment of OFCs	SVM, LR, NB, RF, KNN, DT, and DNN
Dai, et al^[15]	2024	Epigenetic sequencing data from PCW4-10	Pathogenic variant prioritization for OFCs	CNN
Forrest, et al^[16]	2025	Training cohort of 1.3 million individuals with 10 genetic diseases	Genetic variant penetrance estimation	XGBoost
Kamitaki, et al^[17]	2026	Genomic and oral microbiome data from 12 519 individuals	Genetic variant-oral microbiome association analysis	PCA
Guo, et al^[18]	2025	scRNA-seq data of mouse molars from embryonic stages to postnatal days	Cell subtype identification	PCA, UMAP, etc.
Huang, et al^[19]	2025	scRNA-seq data of mouse palate during embryonic development	Cell subtype identification	PCA, UMAP, etc.
Yankee, et al^[20]	2023	scRNA-seq data of human embryonic craniofacial tissues	Cell subtype identification	PCA, UMAP, etc.
Liu, et al^[21]	2025	Paired bulk transcriptomic and single-cell transcriptomic data	Single-cell proteome prediction	Transformer

Reference	Year	Dataset	Objective	AI model
Zhang, et al^[11]	2018	43 GWAS-significant SNPs associated with OFCs	Genetic risk assessment of OFCs	SVM, LR, NB, RF, KNN, DT, and DNN
Dai, et al^[15]	2024	Epigenetic sequencing data from PCW4-10	Pathogenic variant prioritization for OFCs	CNN
Forrest, et al^[16]	2025	Training cohort of 1.3 million individuals with 10 genetic diseases	Genetic variant penetrance estimation	XGBoost
Kamitaki, et al^[17]	2026	Genomic and oral microbiome data from 12 519 individuals	Genetic variant-oral microbiome association analysis	PCA
Guo, et al^[18]	2025	scRNA-seq data of mouse molars from embryonic stages to postnatal days	Cell subtype identification	PCA, UMAP, etc.
Huang, et al^[19]	2025	scRNA-seq data of mouse palate during embryonic development	Cell subtype identification	PCA, UMAP, etc.
Yankee, et al^[20]	2023	scRNA-seq data of human embryonic craniofacial tissues	Cell subtype identification	PCA, UMAP, etc.
Liu, et al^[21]	2025	Paired bulk transcriptomic and single-cell transcriptomic data	Single-cell proteome prediction	Transformer

Reference	Year	Dataset	Objective	AI model
Lee, et al^[23]	2018	3 000 periapical radiographs	Dental caries detection	CNN
Hiraiwa, et al^[24]	2019	CBCT images and panoramic radiographs of mandibular first molars from 400 patients	Root canal morphology classification	CNN
Ekert, et al^[25]	2019	2 001 panoramic radiographs	Periapical lesion diagnosis	DNN
Duman, et al^[26]	2023	220 panoramic radiographs of children aged 6‒9 years	Hyperdontia diagnosis	CNN
Incerti Parenti, et al^[27]	2025	A retrospective dataset of 250 panoramic radiographs from patients aged 6‒13 years	Primary and permanent tooth detection and segmentation	CNN
Sheng, et al^[28]	2023	100 panoramic radiographs	Tooth segmentation	Transformer
Han, et al^[30]	2023	203 salivary gland pathology slides	Segmentation of lymphocytic infiltration areas	DNN
Chiesa-Estomba, et al^[31]	2021	Multi-source clinical data from 345 patients undergoing salivary gland tumor surgery	Prediction of postoperative facial nerve palsy	KNN, RF, BC, and LDC
Jurek, et al^[32]	2020	Fetal ultrasound images	Fetal cleft palate diagnosis	NLP
Shafi, et al^[33]	2020	1 000 prenatal survey records	Risk prediction of cleft lip and palate	DNN
Kuwada, et al^[34]	2021	Panoramic radiographs of 383 patients with cleft conditions (with or without alveolar cleft) and 210 healthy controls	Cleft palate classification	DNN
Zhang, et al^[35]	2020	CBCT images of 21 patients with unilateral or bilateral CLP	Alveolar bone defect volume estimation	CNN
Pactas, et al^[36]	2019	Pre- and post-treatment photographs of 146 orthognathic patients	Assessment of orthognathic surgical outcome	CNN
Huang, et al^[38]	2025	Pre- and post-treatment CBCT images of 100 orthognathic surgery patients	Prediction of postoperative facial appearance	CNN
Adel, et al^[39]	2025	3D facial images of 130 patients	Facial symmetry assessment	DNN
Chen, et al^[40]	2025	CT images of 15 patients	Soft tissue defect assessment	CNN
Saranya, et al^[41]	2022	Bone CT images	FD lesion segmentation	R-CNN
Li, et al^[42]	2025	CT images of 148 FD patients	Prediction of FD lesion growth status	CNN
Zhang, et al^[43]	2024	CT images of 220 patients with FD or OF	Differential diagnosis between FD and OF	RF, SVM, LightGBM, and XGBoost

Reference	Year	Dataset	Objective	AI model
Lee, et al^[23]	2018	3 000 periapical radiographs	Dental caries detection	CNN
Hiraiwa, et al^[24]	2019	CBCT images and panoramic radiographs of mandibular first molars from 400 patients	Root canal morphology classification	CNN
Ekert, et al^[25]	2019	2 001 panoramic radiographs	Periapical lesion diagnosis	DNN
Duman, et al^[26]	2023	220 panoramic radiographs of children aged 6‒9 years	Hyperdontia diagnosis	CNN
Incerti Parenti, et al^[27]	2025	A retrospective dataset of 250 panoramic radiographs from patients aged 6‒13 years	Primary and permanent tooth detection and segmentation	CNN
Sheng, et al^[28]	2023	100 panoramic radiographs	Tooth segmentation	Transformer
Han, et al^[30]	2023	203 salivary gland pathology slides	Segmentation of lymphocytic infiltration areas	DNN
Chiesa-Estomba, et al^[31]	2021	Multi-source clinical data from 345 patients undergoing salivary gland tumor surgery	Prediction of postoperative facial nerve palsy	KNN, RF, BC, and LDC
Jurek, et al^[32]	2020	Fetal ultrasound images	Fetal cleft palate diagnosis	NLP
Shafi, et al^[33]	2020	1 000 prenatal survey records	Risk prediction of cleft lip and palate	DNN
Kuwada, et al^[34]	2021	Panoramic radiographs of 383 patients with cleft conditions (with or without alveolar cleft) and 210 healthy controls	Cleft palate classification	DNN
Zhang, et al^[35]	2020	CBCT images of 21 patients with unilateral or bilateral CLP	Alveolar bone defect volume estimation	CNN
Pactas, et al^[36]	2019	Pre- and post-treatment photographs of 146 orthognathic patients	Assessment of orthognathic surgical outcome	CNN
Huang, et al^[38]	2025	Pre- and post-treatment CBCT images of 100 orthognathic surgery patients	Prediction of postoperative facial appearance	CNN
Adel, et al^[39]	2025	3D facial images of 130 patients	Facial symmetry assessment	DNN
Chen, et al^[40]	2025	CT images of 15 patients	Soft tissue defect assessment	CNN
Saranya, et al^[41]	2022	Bone CT images	FD lesion segmentation	R-CNN
Li, et al^[42]	2025	CT images of 148 FD patients	Prediction of FD lesion growth status	CNN
Zhang, et al^[43]	2024	CT images of 220 patients with FD or OF	Differential diagnosis between FD and OF	RF, SVM, LightGBM, and XGBoost