Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

MILGDF: A Multi-Task Instance-Level Supervised Learning Framework for Oral Cancer Incorporating Local-Global

Chenxi Li^1,2,3, Yan Chen⁴, Lianghui Xu⁵

¹Department of Oral and Maxillofacial Oncology & Surgery, School/Hospital of Stomatology, The First Affiliated Hospital of Xinjiang Medical University, National Clinical Medical Research Institute, Stomatological Research Institute of Xinjiang Uygur Autonomous Region, Urumqi, China.

|June 6, 2026

Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

MILGDF: a multi-task, instance-level supervised model for oral squamous cell carcinoma integrating local-global attention and dynamic decision fusion.

Journal of clinical pathology·2026

Same author

A newly discovered PAX7 variant associated with craniofacial duplication: a case report.

BMC pregnancy and childbirth·2026

Same author

Association of Porphyromonas gingivalis-infected oral squamous cell carcinoma cell-secreted exosomal miR-3648-1-p5 with tumor progression.

Cancer cell international·2026

Same author

Information-Guided Fusion of Multimodal Vibrational Spectroscopy for Disease Diagnosis Based on Symbiotic Attention Decoupled Contrastive Learning.

Analytical chemistry·2026

Same author

The Clinical Effectiveness Comparison Between Modified Crestal Approach and Lateral Window Sinus Floor Elevation: A Prospective Study.

Journal of oral rehabilitation·2026

Same author

Duplicated, Translocated Upper Lip and Maxilla: An Extremely Rare Congenital Craniofacial Anomaly With Novel Genetic Findings.

Journal of oral pathology & medicine : official publication of the International Association of Oral Pathologists and the American Academy of Oral Pathology·2026

Same journal

Authors' Reply "Synchronous Soft-Tissue Lesions in the Oral Cavity in an Adolescent".

Oral diseases·2026

Same journal

Neutrophils From Aged Individuals Release NETs to Impair Oral Wound Healing Through NLRP3 Activation.

Oral diseases·2026

Same journal

Comment on "Multimodal Deep Learning-Based Screening of Degenerative Temporomandibular Joint Disease Using 2D Radiography: A Cost-Effective and Low-Radiation Approach".

Oral diseases·2026

Same journal

Oral Lesions and Dental Status in Individuals Using Oral Tobacco-Free Nicotine Pouches.

Oral diseases·2026

Same journal

Comment on "Synchronous Soft-Tissue Lesions in the Oral Cavity in an Adolescent".

Oral diseases·2026

Same journal

Atypical Whitish Gingival Plaque.

Oral diseases·2026

See all related articles

This summary is machine-generated.

This study introduces the MILGDF model for diagnosing oral squamous cell carcinoma (OSCC) using whole-slide imaging (WSI). The model accurately diagnoses and stages OSCC, outperforming existing methods.

Area of Science:

Digital Pathology
Artificial Intelligence in Medicine
Oncology

Background:

Oral squamous cell carcinoma (OSCC) presents diagnostic challenges due to complex histology.
Whole-slide imaging (WSI) offers potential for improved OSCC diagnosis.
Current diagnostic methods struggle with OSCC's morphological diversity.

Purpose of the Study:

To develop an innovative diagnostic strategy for OSCC using WSI.
To overcome limitations in current OSCC diagnostic approaches.
To enhance the accuracy of OSCC staging and diagnosis.

Main Methods:

Proposed a multi-task learning architecture (MILGDF) integrating local-global attention and dynamic weighted fusion.
Employed instance-level category-specific attention for superior feature extraction.

Keywords:

MILGDF early diagnosis oral squamous cell carcinoma whole‐slide imaging

Related Experiment Videos

Incorporated an adaptive weighting system for optimal performance across prediction tasks.

Main Results:

The MILGDF framework achieved high predictive performance on HIDOC (AUC: 0.952, accuracy: 0.909) and TCGA-OSCC (AUC: 0.745, accuracy: 0.725) datasets.
Demonstrated statistically significant superiority over comparative models in staging and diagnosis.
Validated the model's effectiveness on independent OSCC datasets.

Conclusions:

The MILGDF model accurately diagnoses and stages OSCC using WSI data.
MILGDF performance exceeds existing methods, showing significant clinical potential.
This AI-driven approach offers a promising tool for oral cancer diagnostics.