Medical visual question answering based on question-type reasoning and semantic space constraint
作者：
Highlights：
• A new Framework has been proposed for the medical visual question answering tasks.
• The question-type reasoning module was proposed for question type classification.
• A joint semantic space was structured for rich answer representation.
• A composition loss function was designed for the Med-VQA model.
• The results reveal that new framework have good contribution to medical visual question answering tasks.
摘要
•A new Framework has been proposed for the medical visual question answering tasks.•The question-type reasoning module was proposed for question type classification.•A joint semantic space was structured for rich answer representation.•A composition loss function was designed for the Med-VQA model.•The results reveal that new framework have good contribution to medical visual question answering tasks.
论文关键词：Medical visual question answering,Question-type reasoning,Semantic space constraint,Attention mechanism
论文评审过程：Received 18 October 2021, Revised 14 May 2022, Accepted 27 June 2022, Available online 30 June 2022, Version of Record 6 July 2022.
论文官网地址：https://doi.org/10.1016/j.artmed.2022.102346
