Deep learning approach to identify histological features associated with lymph node metastasis following primary tumor excision in patients with tongue squamous cell carcinoma

Key Points

• The aim is to determine if deep learning can identify features in tissue samples that correlate with lymph node metastasis in tongue squamous cell carcinoma.
• Analyzed routine hematoxylin-eosin stained tissue sections
• Utilized deep learning algorithms for image analysis
• Focus on clinically node-negative early-stage disease
• Deep learning effectively identifies histological features linked to lymph node metastasis
• Supports postoperative risk stratification
• Highlights potential in early-stage disease assessment