What type of study is this?

This is a Systematic Review study.

September 23, 2025Open Access

Can Large Language Models Reliably Interpret Radiology Reports? A Systematic Evaluation for Tumor Progression Classification

View Full Paper

Ask AI

Bookmark

Authors

VPValentin PohyerInserm CMConstance de Margerie‐MellonInserm LPLaetitia PerronneNorthwestern University

Discussion

Member takes

Overview

Systematic evaluation reveals large language models outperform BERT in classifying tumor progression, indicating significant potential for clinical data processing.

Key Points

Large language models demonstrated superior performance in classifying tumor progression from radiology reports, with 310 cases analyzed.
Models were tested on different architectures and prompting strategies, achieving better results than traditional rule-based and BERT-based methods.
The evaluation highlighted a trade-off between computational resources and the need for human annotations, affecting cost and development time.
Selected models showed better efficiency than fine-tuned BERT, suggesting paths for optimizing the use of clinical text data.

Cite This Study

Pohyer et al. (2025) studied this question.

synapsesocial.com/papers/68d4759931b076d99fa6da91 https://doi.org/10.21203/rs.3.rs-7630430/v1

View Full Paper

Ask AI

Bookmark

September 23, 2025Open Access

Can Large Language Models Reliably Interpret Radiology Reports? A Systematic Evaluation for Tumor Progression Classification

View Full Paper

Ask AI

Bookmark

Authors

VPValentin PohyerInserm CMConstance de Margerie‐MellonInserm LPLaetitia PerronneNorthwestern University

Discussion

Member takes

Overview

Systematic evaluation reveals large language models outperform BERT in classifying tumor progression, indicating significant potential for clinical data processing.

Key Points

Large language models demonstrated superior performance in classifying tumor progression from radiology reports, with 310 cases analyzed.
Models were tested on different architectures and prompting strategies, achieving better results than traditional rule-based and BERT-based methods.
The evaluation highlighted a trade-off between computational resources and the need for human annotations, affecting cost and development time.
Selected models showed better efficiency than fine-tuned BERT, suggesting paths for optimizing the use of clinical text data.

Cite This Study

Pohyer et al. (2025) studied this question.

synapsesocial.com/papers/68d4759931b076d99fa6da91 https://doi.org/10.21203/rs.3.rs-7630430/v1

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark