EUVIMEDEUVIMED entdecken
Crossref

MultiMedBench: A Scenario-Aware Benchmark for Evaluating Knowledge Editing in Medical VQA

Shengtao Wen, Haodong Chen, Yadong Wang, Zhongying Pan, Xiang Chen, Yu Tian, Bo Qian, Dong Liang, Sheng-Jun Huang

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (40) · 33872-33880

DOI: 10.1609/aaai.v40i40.40679 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Shengtao Wen und 8 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Knowledge editing (KE) provides a scalable approach for updating factual knowledge in large language models without full retraining. While previous studies have demonstrated effectiveness in general domains and medical QA tasks, little attention has been paid to KE in multimodal medical scenarios. Unlike text-only settings, medical KE demands integrating updated knowledge with visual reasoning to support safe and interpretable clinical decisions. To address this gap, we propose MultiMedBench, the first benchmark tailored to evaluating KE in clinical multimodal tasks. Our framework spans both understanding and reasoning task types, defines a three-dimensional metric suite (reliability, generality, and locality), and supports cross-paradigm comparisons across general and domain-specific models. We conduct extensive experiments under single-editing and lifelong-editing settings. Results suggest that current methods struggle with generalization and long-tail reasoning, particularly in complex clinical workflows. We further present an efficiency analysis (e.g., edit latency, memory footprint), revealing practical trade-offs in real-world deployment across KE paradigms. Overall, MultiMedBench not only reveals the limitations of current approaches but also provides a solid foundation for developing clinically robust knowledge editing techniques in the future.

Zitieren

APA

Wen, S., Chen, H., Wang, Y., Pan, Z., Chen, X., Tian, Y., Qian, B., Liang, D., & Huang, S. (2026). MultiMedBench: A Scenario-Aware Benchmark for Evaluating Knowledge Editing in Medical VQA. Proceedings of the AAAI Conference on Artificial Intelligence, 40(40), 33872-33880. https://doi.org/10.1609/aaai.v40i40.40679

Vancouver

Wen S, Chen H, Wang Y, Pan Z, Chen X, Tian Y, et al. MultiMedBench: A Scenario-Aware Benchmark for Evaluating Knowledge Editing in Medical VQA. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(40):33872-33880. doi:10.1609/aaai.v40i40.40679

BibTeX
@article{Wen2026,
  author = {Wen, Shengtao and Chen, Haodong and Wang, Yadong and Pan, Zhongying and Chen, Xiang and Tian, Yu and Qian, Bo and Liang, Dong and Huang, Sheng-Jun},
  title = {MultiMedBench: A Scenario-Aware Benchmark for Evaluating Knowledge Editing in Medical VQA},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {40},
  pages = {33872--33880},
  doi = {10.1609/aaai.v40i40.40679}
}
Kostenlos starten →