Scholarlink

Medical imaging is an integral part of modern healthcare, aiding in the diagnosis of various abnormalities. However, challenges such as diagnostic errors remain a significant concern. This research introduces a multimodal large language model designed to generate comprehensive clinical summaries directly from radiology images. The proposed system leverages the SLAKE dataset, which includes imaging modalities such as X-rays, MRIs, and CT scans, to extract critical features, detect abnormalities, and generate precise textual reports. By integrating state-of-the-art image analysis techniques with natural language processing, the model ensures accurate, consistent, and reliable outputs that align with medical terminology. This solution is particularly valuable for streamlining diagnostic workflows and enhancing the reliability of clinical decision-making. The approach highlights the potential of combining vision and language capabilities to create robust, scalable systems that contribute to improving diagnostic accuracy and supporting healthcare providers in delivering better patient outcomes.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Radiology Images to Text Synthesis Using Multimodal Large Language Models for Generating Clinical Summaries

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Akshay Poojary,
  • Varsha Sajjanavar,
  • N. Abhishek,
  • Tarun Ejanthkar,
  • Prashant Narayankar

摘要

Medical imaging is an integral part of modern healthcare, aiding in the diagnosis of various abnormalities. However, challenges such as diagnostic errors remain a significant concern. This research introduces a multimodal large language model designed to generate comprehensive clinical summaries directly from radiology images. The proposed system leverages the SLAKE dataset, which includes imaging modalities such as X-rays, MRIs, and CT scans, to extract critical features, detect abnormalities, and generate precise textual reports. By integrating state-of-the-art image analysis techniques with natural language processing, the model ensures accurate, consistent, and reliable outputs that align with medical terminology. This solution is particularly valuable for streamlining diagnostic workflows and enhancing the reliability of clinical decision-making. The approach highlights the potential of combining vision and language capabilities to create robust, scalable systems that contribute to improving diagnostic accuracy and supporting healthcare providers in delivering better patient outcomes.