What does this research mean for the field?

A combination approach utilizing specific pre-trained language models (JMedRoBERTa for label T, Tohoku-BERT-v3 for label N, and UTH-BERT for label M) achieves superior accuracy in classifying radiology reports for TNM staging. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

The aim was to classify radiology reports based on TNM staging criteria using various language models.

April 1, 2026Open Access

NAISTSOCRR at the NTCIR-17 MedNLP-SC Radiology Report Subtask

Key Points

The aim was to classify radiology reports based on TNM staging criteria using various language models.
Participated in NTCIR17's RR-TNM Subtask
Utilized three pre-trained language models: JMedRoBERTa, Tohoku-BERT-v3, and UTH-BERT
Employed a combination approach for improved classification
Achieved the highest accuracy of 0.3704 among all participants
Effectiveness of the strategy was emphasized through method differentiation

Abstract

This paper describes how we tackled the Medical Natural Language Processing for Radiology Report TNM staging (RR-TNM) Subtask as participants of NTCIR17. The RR-TNM Subtask is a MedNLP-SC original task to classify radiology reports under multiple criteria. We introduced three different methods based on pre-trained language models (PLMs), including a medical-specific model. Notably, our combination approach, utilizing JMedRoBERTa (manbyo-wordpiece) for label T, Tohoku-BERT-v3 for label N, and UTH-BERT for label M, achieved an accuracy of 0.3704 on the test data. This performance was the highest among all participants, emphasizing the effectiveness of our strategy.

Read Full Paperexternally

AIに質問

Bookmark

View Full Paper