CBM-RAG: Demonstrating Enhanced Interpretability in Radiology Report Generation with Multi-Agent RAG and Concept Bottleneck Models
Open accessThe DOI matches its Crossref record; 14 of the 14 references it cites were found in a registry or an index.
Abstract
Advancements in generative Artificial Intelligence (AI) hold great promise for automating radiology workflows, yet challenges in interpretability and reliability hinder clinical adoption. This paper presents an automated radiology report generation framework that combines Concept Bottleneck Models (CBMs) with a Multi-Agent Retrieval-Augmented Generation (RAG) system to bridge AI performance with clinical explainability. CBMs map chest X-ray features to human-understandable clinical concepts, enabling transparent disease classification. Meanwhile, the RAG system integrates multi-agent collaboration and external knowledge to produce contextually rich, evidence-based reports. Our demonstration showcases the system's ability to deliver interpretable predictions, mitigate hallucinations, and generate high-quality, tailored reports with an interactive interface addressing accuracy, trust, and usability challenges. This framework provides a pathway to improving diagnostic consistency and empowering radiologists with actionable insights.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Consistent with this distribution, post hoc explainability was not used as a standalone strategy but appeared exclusively in hybrid systems that combined intrinsic agent reasoning with attribution-based methods [20,22].”
From Explainable Agentic Artificial Intelligence in Healthcare: A Scoping Review (2026) · cites this paper as [20]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“These systems also reported improvements in trust-related metrics, including reduced hallucinations and increased perceived reliability of generated reports [20,23].”
From Explainable Agentic Artificial Intelligence in Healthcare: A Scoping Review (2026) · cites this paper as [20]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Nine studies met the inclusion criteria.”
From Explainable Agentic Artificial Intelligence in Healthcare: A Scoping Review (2026) · cites this paper as [20]
The cited source says what the claim attributes to it.
In the citing paper
“Study designs varied, with several studies representing proof-of-concept implementations (e.g., CBM-RAG, ENTAgents, ArgMed-AgentN, DSM5AgentFlow, and BRAD) [20,21,24,25,27].”
From Explainable Agentic Artificial Intelligence in Healthcare: A Scoping Review (2026) · cites this paper as [20]
The cited source says what the claim attributes to it.
In the citing paper
“The primary clinical and biomedical tasks addressed included clinical decision support systems (CDSSs) [20,21,22,24,27], question answering (QA) [27], mental health screening and assessment [24], diagnostic and report generation or summarization [20,22,23], drug–target interaction prediction [19], biomarker discovery [25], and single-cell annotation [26].”
From Explainable Agentic Artificial Intelligence in Healthcare: A Scoping Review (2026) · cites this paper as [20]
1 more claim about this paper
Create a free account to see every claim other papers make about this one.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Can AI help in screening viral and COVID-19 pneumonia?
Chowdhury, M. E. H., Rahman, T., Khandakar, A., Mazhar, R., Kadir, M. A., Mahbub, Z. B., Islam, K. R., Khan, M. S., Iqbal, A., Emadi, N. A., et al. (2020) · IEEE Access · doi:10.1109/access.2020.3010287 · Cited by 1,913
Self-citationMedCLIP: Contrastive Learning from Unpaired Medical Images and Text
Wang, Z., Wu, Z., Agarwal, D., & Sun, J. (2022) · Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2022.emnlp-main.256 · Cited by 761
ReAct: Synergizing Reasoning and Acting in Language Models
Yao, S., Zhao, J., Yu, D., Du, N., Shafran, I., Narasimhan, K., & Cao, Y. (2022) · arXiv · doi:10.48550/arxiv.2210.03629 · Cited by 608
Label-Free Concept Bottleneck ModelsPreprint
Oikarinen, T., Das, S., Nguyen, L. M., & Weng, T.-W. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2304.06129 · Cited by 463
Jiang, A. Q., Sablayrolles, A., Mensch, A., Bamford, C., Chaplot, D. S., Casas, D. D. L., Bressand, F., Lengyel, G., Lample, G., Saulnier, L., et al. (2023) · arXiv · doi:10.48550/arxiv.2310.06825 · Cited by 340
A Multimodal Biomedical Foundation Model Trained from Fifteen Million Image–Text Pairs
Zhang, S., Xu, Y., Usuyama, N., Xu, H., Bagga, J., Tinn, R., Preston, S., Rao, R., Wei, M., Valluri, N., et al. (2025) · NEJM AI · doi:10.1056/aioa2400640 · Cited by 283
R2GenGPT: Radiology Report Generation with frozen LLMs
Wang, Z., Liu, L., Wang, L., & Zhou, L. (2023) · Meta-Radiology · doi:10.1016/j.metrad.2023.100033 · Cited by 192
Hallucinations in LLMs: Understanding and Addressing Challenges
Perković, G., Drobnjak, A., & Botički, I. (2024) · 2024 47th MIPRO ICT and Electronics Convention (MIPRO) · doi:10.1109/mipro60963.2024.10569238 · Cited by 176
Zhang, S., Xu, Y., Usuyama, N., Xu, H., Bagga, J., Tinn, R., Preston, S., Rao, R., Wei, M., Valluri, N., et al. (2023) · arXiv · doi:10.48550/arxiv.2303.00915 · Cited by 101
Robust and Interpretable Medical Image Classifiers via Concept Bottleneck ModelsPreprint
Yan, A., Wang, Y., Zhong, Y., He, Z., Karypis, P., Wang, Z., Dong, C., Gentili, A., Hsu, C., Shang, J., et al. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2310.03182 · Cited by 62
Research works printed but not linked: 6
- Zhihong Chen, Maya Varma, Jean-Benoit Delbrouck, Magdalini Paschali, Louis Blankemeier, Dave Van Veen, Jeya Maria Jose Valanarasu, Alaa Youssef, Joseph Paul Cohen, Eduardo Pontes Reis, Emily Tsai, Andrew Johnston, Cameron Olsen, Tanishq Mathew Abraham, Sergios Gatidis, Akshay S Chaudhari, and Curtis Langlotz. 2024. CheXagent: Towards a Foundation Model for Chest X-Ray Interpretation. In AAAI 2024 Spring Symposium on Clinical Foundation Models. https://openreview.net/forum?id=P3LOmrZWGR
- Christopher Ifeanyi Eke and Liyana Shuib. 2024. The role of explainability and transparency in fostering trust in AI healthcare systems: a systematic literature review open issues and potential solutions. Neural Computing and Applications (2024) 1–36.
- Pang Wei Koh, Thao Nguyen, Yew Siang Tang, Stephen Mussmann, Emma Pierson, Been Kim, and Percy Liang. 2020. Concept bottleneck models. In International conference on machine learning. PMLR, 5338–5348.
- Patrick Lewis Ethan Perez Aleksandra Piktus Fabio Petroni Vladimir Karpukhin Naman Goyal Heinrich Küttler Mike Lewis Wen-tau Yih Tim Rocktäschel et al. 2020. Retrieval-augmented generation for knowledge-intensive nlp tasks. Advances in Neural Information Processing Systems 33 (2020) 9459–9474.
- Zhengliang Liu, Aoxiao Zhong, Yiwei Li, Longtao Yang, Chao Ju, Zihao Wu, Chong Ma, Peng Shu, Cheng Chen, Sekeun Kim, et al. 2023. Tailoring large language models to radiology: A preliminary approach to llm adaptation for a highly specialized domain. In International Workshop on Machine Learning in Medical Imaging. Springer, 464–473.
- Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, and Ilya Sutskever. 2023. Robust speech recognition via large-scale weak supervision. In International conference on machine learning. PMLR, 28492–28518.
Cited by
Tripathi, S., Cook, T. S., & Kim, W. (2026) · Radiology · doi:10.1148/radiol.252730 · Cited by 23
Generative AI for healthcare: fundamentals, challenges, and perspectives
Chen, G., Liu, C., Anne, O. G., Tan, M., Xie, Z., Yin, J., Yip, J. W. L., Zhang, W., ZHU, J., & Ooi, B. C. (2026) · Science China Information Sciences · doi:10.1007/s11432-025-4927-2 · Cited by 10
Explainable Agentic Artificial Intelligence in Healthcare: A Scoping Review
Collaco, B. G., Prabha, S., Gomez-Cabello, C. A., Haider, S. A., Genovese, A., Wood, N. G., Gopala, N., Raman, R., Hester, E. O., & Forte, A. J. (2026) · Bioengineering · doi:10.3390/bioengineering13050513 · Cited by 3
“For example, CBM-RAG separated disease-specific reasoning from report synthesis [20]; ADAM-1 distinguished computational analysis from narrative interpretation [22]; and CASSIA partitioned annotation, validation, quality scoring, and reporting into discrete stages [26].”
InfluentialHamma, M. B., Merrouni, Z. A., Frikh, B., & Ouhbi, B. (2026) · Artificial Intelligence Review · doi:10.1007/s10462-026-11556-z · Cited by 2
TRACE: A Concept Bottleneck Model for Longitudinal 3D Glioblastoma Response AssessmentPreprint
Tarek, A., Saberr, H., Elghonemy, H., Afify, Y., Basha, T., Bhatti, O. S., Selim, A. M., Alam, H., & Sonntag, D. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2606.30313 · Cited by 2
Self-citation
Metadata
- Authors
- Alam, H. M. T., Srivastav, D., Mohamed Selim, A., Kadir, M. A., Shuvo, M. M. H., & Sonntag, D.
- Affiliations
- German Research Centre for Artificial Intelligence (DE)
- German Research Center for Artificial Intelligence (DFKI), Saarbrücken, Germany
- Carl von Ossietzky Universität Oldenburg (DE)
- Dhaka Medical College and Hospital (BD)
- ORCID
- Venue
- Companion Proceedings of the 17th ACM SIGCHI Symposium on Engineering Interactive Computing Systems
- Publisher
- ACM
- Published
- 2025-06-22
- Pages
- 59-61
- OpenAlex
- W4411494168
- Cited by
- 4
- References
- 7
- Citations by year
- 2026: 4
- FWCI
- 4.18
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- cc-by-nc-sa
- Locations
- ACM SIGCHI Symposium on Engineering Interactive Computing Systems (EICS) (publishedVersion)
- arXiv (Cornell University) (submittedVersion, cc-by-nc-sa)
- arXiv (Cornell University)
- Funders
- Bundesministerium für Bildung und Forschung (01IW23002)
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Multimodal Machine Learning Applications
- Keywords
- radiology report generation, retrieval-augmented generation, multi-agent systems, concept bottleneck models, chest X-ray classification, interpretability, hallucination mitigation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: CBM-RAG: Demonstrating Enhanced Interpretability in Radiology Report Generation with Multi-Agent RAG and Concept Bottleneck Models. https://app.cochat.ai/papers/10.1145/3731406.3731970
- The paper
- Alam, H. M. T., Srivastav, D., Mohamed Selim, A., Kadir, M. A., Shuvo, M. M. H., & Sonntag, D. (2025). CBM-RAG: Demonstrating Enhanced Interpretability in Radiology Report Generation with Multi-Agent RAG and Concept Bottleneck Models. Companion Proceedings of the 17th ACM SIGCHI Symposium on Engineering Interactive Computing Systems. https://doi.org/10.1145/3731406.3731970
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.