SICAR at RRG2024: GPU Poor’s Guide to Radiology Report Generation
Issued Date
2024-01-01
Resource Type
Scopus ID
2-s2.0-85204496103
Journal Title
BioNLP 2024 - 23rd Meeting of the ACL Special Interest Group on Biomedical Natural Language Processing, Proceedings of the Workshop and Shared Tasks
Start Page
635
End Page
644
Rights Holder(s)
SCOPUS
Bibliographic Citation
BioNLP 2024 - 23rd Meeting of the ACL Special Interest Group on Biomedical Natural Language Processing, Proceedings of the Workshop and Shared Tasks (2024) , 635-644
Suggested Citation
Udomlapsakul K., Pengpun P., Saengja T., Veerakanjana K., Tiankanon K., Khlaisamniang P., Supholkhan P., Chinkamol A., Aussavavirojekul P., Phimsiri H., Sripo T., Boonnag C., Tongdee T., Siriapisith T., Saiviroonporn P., Kinchagawat J., Ittichaiwong P. SICAR at RRG2024: GPU Poor’s Guide to Radiology Report Generation. BioNLP 2024 - 23rd Meeting of the ACL Special Interest Group on Biomedical Natural Language Processing, Proceedings of the Workshop and Shared Tasks (2024) , 635-644. 644. Retrieved from: https://repository.li.mahidol.ac.th/handle/20.500.14594/101397
Title
SICAR at RRG2024: GPU Poor’s Guide to Radiology Report Generation
Corresponding Author(s)
Other Contributor(s)
Abstract
Radiology report generation (RRG) aims to create free-text radiology reports from clinical imaging. Our solution employs a lightweight multimodal language model (MLLM) enhanced with a two-stage post-processing strategy, utilizing a Large Language Model (LLM) to boost diagnostic accuracy and ensure patient safety. We introduce the "First, Do No Harm" SafetyNet, which incorporates X-Raydar, an advanced X-ray classification model, to cross-verify the model outputs and specifically address false negative errors from the MLLM. This comprehensive approach combines the efficiency of lightweight models with the robustness of thorough post-processing techniques, offering a reliable solution for radiology report generation. Our system achieved fourth place on the F1-Radgraph metric for findings generation in the Radiology Report Generation Shared Task (RRG24).1,..