From Multimodal Data to Clinical Insight: An Explainable Model for Preoperative Salivary Gland Lesion Diagnosis
Oral Diseases · 2026-05 · 32 (5) · 1315-1329
DOI: 10.1111/odi.70199 ↗Kurz gefasst
Fachartikel in der Zeitschrift Oral Diseases, online erschienen am 12.03.2026. Verfasst von Jing Feng und 6 weiteren Autor:innen. Beteiligte Einrichtung: Department of Ultrasound, Shanghai Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China. Hauptthemen laut MEDLINE: Salivary Gland Neoplasms, Adenoma, Pleomorphic, Support Vector Machine. Die Arbeit nennt 31 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: Discipline Group Construction Project of the Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine., Science and Technology Commission of Shanghai Municipality.
Abstract (Originalsprache)
Objective
To develop and validate a multimodal dual‐step support vector machine model (SVM‐DualNet) for the preoperative three‐class classification of salivary gland lesions (SGLs) to support clinical decision‐making.
Methods
We retrospectively collected clinical, conventional ultrasound (CUS), shear wave elastography (SWE), and radiomics features from 284 patients with SGLs. For malignancy discrimination and pleomorphic adenoma identification, linear SVM models based on different modality combinations were constructed and compared. The best‐performing binary models were sequentially combined to form SVM‐DualNet. SHapley Additive exPlanations (SHAP) were applied for global and case‐level interpretation and incorporated into diagnostic assistance. Clinical utility was evaluated by comparing the junior radiologist's diagnostic performance before and after SHAP assistance and by comparison with the senior radiologist.
Results
In the test cohort, SVM‐DualNet achieved a balanced accuracy of 0.76 and a macro F1 score of 0.82 for the three‐class classification. The binary models discriminated malignancy and pleomorphic adenoma with AUCs of 0.90 (95% CI: 0.82–0.97) and 0.85 (95% CI: 0.76–0.94), respectively. SHAP‐assisted review improved the junior radiologist's balanced accuracy from 0.55 to 0.70 and macro F1 from 0.57 to 0.75, approaching the senior radiologist's performance.
Conclusions
The model provides reliable preoperative classification of SGLs and can assist clinicians in decision‐making.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Ultrasound, Shanghai Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China.
- Computer Science, Clark University, Worcester, Massachusetts, USA.
ORCID-Profile: Ang Ang Ding, Ping Xiong
Förderung
- Discipline Group Construction Project of the Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine.
- Science and Technology Commission of Shanghai Municipality
Zitieren
APA
Feng, J., Yan, Y., Li, J., Liu, S. L., Ding, A. A., He, C., & Xiong, P. (2026). From Multimodal Data to Clinical Insight: An Explainable Model for Preoperative Salivary Gland Lesion Diagnosis. Oral Diseases, 32(5), 1315-1329. https://doi.org/10.1111/odi.70199
Vancouver
Feng J, Yan Y, Li J, Liu SL, Ding AA, He C, et al. From Multimodal Data to Clinical Insight: An Explainable Model for Preoperative Salivary Gland Lesion Diagnosis. Oral Diseases. 2026;32(5):1315-1329. doi:10.1111/odi.70199
BibTeX
@article{Feng2026,
author = {Feng, Jing and Yan, Yuqi and Li, Jia and Liu, Sai Liang and Ding, Ang Ang and He, Chang and Xiong, Ping},
title = {From Multimodal Data to Clinical Insight: An Explainable Model for Preoperative Salivary Gland Lesion Diagnosis},
journal = {Oral Diseases},
year = {2026},
volume = {32},
number = {5},
pages = {1315--1329},
doi = {10.1111/odi.70199}
}
EUVIMED