おめでとうございます！
ラシド研の研究生（2025年4月からD1）のエルボラディ・アハメドさんがIFMIA2025でBest Poster Awardを受賞しました。
Research student Ahmed Elboardy (D1 from Apr. 2025) received the Best Poster Award from the International Forum on Medical Imaging in Asia (IFMIA2025)
Summary: Automatic generation of medical reports is a challenging task that requires significant time and professional skills. In the ear of aging society, it is important to develop a trusted technology that enable automatic generation of diagnosis reports directly from medical images. Vision language models (VLMs) provide a useful tool for healthcare data analysis. However, it is unclear how different models performs in solving different clinical problems. In this study, we evaluated the performance of various VLMs in radiology report generation models by contrasting their outputs with expert-revised reports. Results indicate that some VLMs can performs at level close to human experts considering the problem of brain cancer diagnosis from multisequence MRI.
Acknowledgment: This work was supported by JST PRESTO Grant Number JPMJPR23P7, Japan
