Enhancing medical Q&A systems with multimodal knowledge graphs and dual-layer attention mechanisms
Guoqiang Qiu, Qingni Yuan, Yi Wang, Pengju Qu, Wei Jia · 2026
Read Enhancing medical Q&A systems with multimodal knowledge graphs and dual-layer attention mechanisms by Guoqiang Qiu, Qingni Yuan, Yi Wang, Pengju Qu, Wei Jia (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0353112.txt.
Description
Guoqiang Qiu and four co-authors present this 2026 PLOS One paper on a medical question-answering system. The system is built to grasp what patients actually mean. A knowledge graph links medical terms with more than fifteen thousand medical images. A two-layer attention model sorts out the intent behind multi-part questions. A large language model picks out medical terms with no hand-labeled training data.
In benchmark tests the intent model topped several rival systems. The term finder recalled medical terms well too. A visual search tool built on the graph let users explore results. Users judged the system effective across varied medical questions. The authors offer the design as a route to medical chat tools. Such tools could handle complex, image-linked queries with less manual data work.
Catalog record
| CREATOR | Guoqiang Qiu, Qingni Yuan, Yi Wang, Pengju Qu, Wei Jia |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 65198 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |