โก Quick Summary
This systematic review evaluated artificial intelligence (AI) models for diagnosing and classifying malignant salivary gland tumors (MSGTs), highlighting their potential in improving diagnostic accuracy. The findings suggest promising performance, particularly in benign-malignant discrimination, but also reveal significant limitations due to high risk of bias and heterogeneity in study designs.
๐ Key Details
- ๐ Dataset: 1,922 participants across eight studies
- ๐งฉ Features used: Radiologic, histopathologic, and molecular data
- โ๏ธ Technology: AI/ML models including radiomics and deep learning
- ๐ Performance: AUCs of 0.890 and 0.745 for benign-malignant discrimination
๐ Key Takeaways
- ๐ MSGTs are rare and biologically diverse tumors, complicating diagnosis.
- ๐ก AI approaches are being explored as adjunctive tools for diagnosis.
- ๐ฉโ๐ฌ Eight studies were included, focusing on various AI/ML models.
- ๐ External validation was limited, with only two studies reporting it.
- ๐ High risk of bias was noted in many studies due to retrospective designs.
- ๐ Three diagnostic domains were identified: malignancy discrimination, subtype classification, and molecular taxonomy.
- ๐ Heterogeneity in model construction and validation strategies hindered meta-analysis.
- ๐ฎ Future research is needed to enhance the robustness of AI models in this field.
๐ Background
Malignant salivary gland tumors (MSGTs) are infrequent yet complex neoplasms that pose significant challenges in terms of diagnosis and classification. The traditional methods often suffer from interobserver variability, making it difficult for clinicians to reach a consensus. The integration of artificial intelligence into this domain offers a promising avenue for enhancing diagnostic accuracy and consistency.
๐๏ธ Study
This systematic review adhered to the PRISMA 2020 guidelines and aimed to critically assess AI/ML models developed for MSGTs. The researchers conducted a comprehensive search across databases such as Embase, PubMed/MEDLINE, and Scopus, focusing on studies that reported extractable performance metrics for AI diagnostic models in human salivary gland tumor cohorts.
๐ Results
Out of 1,265 records screened, only eight studies met the inclusion criteria, encompassing various AI methodologies including CT/MRI radiomics, whole-slide histopathology deep learning, and DNA methylation-based classification. The studies reported AUCs of 0.890 and 0.745 for distinguishing between benign and malignant tumors, indicating a degree of effectiveness in AI applications. However, the high heterogeneity in study designs and validation methods limited the ability to conduct a meta-analysis.
๐ Impact and Implications
The findings from this review underscore the potential of AI/ML models to significantly improve the diagnostic landscape for MSGTs. By refining the processes of malignancy discrimination and histopathologic classification, these technologies could lead to better patient outcomes and more personalized treatment strategies. However, the limitations highlighted in the current evidence base call for further research to establish more robust and reliable AI applications in clinical settings.
๐ฎ Conclusion
This systematic review illustrates the promising role of AI in the diagnosis of malignant salivary gland tumors, particularly in preoperative assessments. While the current models show potential, the high risk of bias and variability in study designs necessitate further investigation to enhance their clinical applicability. The future of AI in oncology looks bright, and continued research will be crucial in unlocking its full potential.
๐ฌ Your comments
What are your thoughts on the integration of AI in diagnosing complex tumors like MSGTs? We would love to hear your insights! ๐ฌ Share your comments below or connect with us on social media:
Artificial-Intelligence-Based Radiologic, Histopathologic, and Molecular Models for the Diagnosis and Classification of Malignant Salivary Gland Tumors: A Systematic Review and Functional Meta-Synthesis.
Abstract
BACKGROUND/OBJECTIVES: Malignant salivary gland tumors (MSGTs) are rare, biologically heterogeneous neoplasms in which histopathologic diagnosis and classification are challenging and subject to interobserver variability. Artificial intelligence (AI) approaches using radiologic, histopathologic, and molecular data, including radiomics, deep learning, and biomarker-based models, have been proposed as adjunctive diagnostic tools. This systematic review aimed to identify and critically appraise AI/ML models across radiologic, histopathologic, and molecular domains for distinct diagnostic tasks in MSGTs, and to integrate their diagnostic roles through a functional meta-synthesis.
METHODS: We conducted a PRISMA 2020-compliant systematic review. Embase, PubMed/MEDLINE, and Scopus were searched from inception to February 2026. Eligible studies developed or validated AI/ML diagnostic or classification models in human salivary gland tumor cohorts and reported extractable performance metrics.
RESULTS: From 1265 records, eight studies (1922 participants) met the inclusion criteria, spanning CT/MRI radiomics or deep learning (n = 4), whole-slide histopathology deep learning (n = 3), and DNA methylation-based classification (n = 1). External validation was reported in two CT-based benign-malignant discrimination studies, with AUCs of 0.890 (95% CI 0.844-0.937) and 0.745 (95% CI 0.699-0.791). Heterogeneity in model construction, outcome definitions, and validation strategies precluded meta-analysis. Risk of bias was frequently high in QUADAS-2/PROBAST assessments, driven by retrospective sampling, limited blinding, and analysis-related concerns, while calibration and utility were rarely assessed.
CONCLUSIONS: AI/ML models for MSGTs demonstrate promising diagnostic performance, particularly for preoperative benign-malignant discrimination, but the current evidence base is limited by heterogeneity, predominantly internal validation, and high risk of bias. The functional meta-synthesis identified three convergent diagnostic domains: malignancy discrimination, histopathologic subtype classification, and molecular/epigenetic taxonomy refinement.
Author: [‘Ardila CM’, ‘Pineda-Vรฉlez E’, ‘Vivares-Builes AM’, ‘Dรญaz-Laclaustra AI’]
Journal: Med Sci (Basel)
Citation: Ardila CM, et al. Artificial-Intelligence-Based Radiologic, Histopathologic, and Molecular Models for the Diagnosis and Classification of Malignant Salivary Gland Tumors: A Systematic Review and Functional Meta-Synthesis. Artificial-Intelligence-Based Radiologic, Histopathologic, and Molecular Models for the Diagnosis and Classification of Malignant Salivary Gland Tumors: A Systematic Review and Functional Meta-Synthesis. 2026; 14:(unknown pages). doi: 10.3390/medsci14020183