Classification of cancer TNM stage from Japanese radiology report using on-premise LLM at NTCIR-17 MedNLP-SC RR-TNM subtask

Key Points

• To classify cancer TNM stages using a language model, enhancing accuracy through data augmentation.
• Created extensive question-and-answer pairs for TNM classification
• Utilized an on-premise language model for analysis
• Participated in the NTCIR-17 MedNLP-SC shared task
• Observed improved accuracy in TNM classification
• Significant improvement noted particularly for the M stage