Image to Label to Answer: An Efficient Framework for Enhanced Clinical Applications in Medical Visual Question Answering

Jianfeng Wang, Kah Phooi Seng*, Yi Shen, Li Minn Ang, Difeng Huang

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Medical Visual Question Answering (Med-VQA) faces significant limitations in application development due to sparse and challenging data acquisition. Existing approaches focus on multi-modal learning to equip models with medical image inference and natural language understanding, but this worsens data scarcity in Med-VQA, hindering clinical application and advancement. This paper proposes the ITLTA framework for Med-VQA, designed based on field requirements. ITLTA combines multi-label learning of medical images with the language understanding and reasoning capabilities of large language models (LLMs) to achieve zero-shot learning, meeting natural language module needs without end-to-end training. This approach reduces deployment costs and training data requirements, allowing LLMs to function as flexible, plug-and-play modules. To enhance multi-label classification accuracy, the framework uses external medical image data for pretraining, integrated with a joint feature and label attention mechanism. This configuration ensures robust performance and applicability, even with limited data. Additionally, the framework clarifies the decision-making process for visual labels and question prompts, enhancing the interpretability of Med-VQA. Validated on the VQA-Med 2019 dataset, our method demonstrates superior effectiveness compared to existing methods, confirming its outstanding performance for enhanced clinical applications.

Original languageEnglish
Article number2273
JournalElectronics (Switzerland)
Volume13
Issue number12
DOIs
Publication statusPublished - Jun 2024

Keywords

  • attention mechanisms
  • large language models (LLMs)
  • medical visual question answering (Med-VQA)
  • multi-label learning
  • zero-shot learning

Cite this