mlsanityData to 7 Oct 2026

Calibrated Confidence Expression for Radiology Report Generation

arXiv 2603.29492 ↗First posted 31 Mar 2026cs.CLcited by 0

On the map: Multimodal LLMs / vision-language models → Medical vision-language models

Abstract (arXiv)

Safe deployment of Large Vision-Language Models (LVLMs) in radiology report generation requires not only accurate predictions but also clinically interpretable indicators of when outputs should be thoroughly reviewed, enabling selective radiologist verification and reducing the risk of hallucinated findings influencing clinical decisions. One intuitive approach to this is verbalized confidence, where the model explicitly states its certainty. However, current state-of-the-art language models are often overconfident, and research on calibration in multimodal settings such as radiology report generation is limited. To address this gap, we introduce ConRad (Confidence Calibration for Radiology Reports), a reinforcement learning framework for fine-tuning medical LVLMs to produce calibrated verbalized confidence estimates alongside radiology reports. We study two settings: a single report-level confidence score and a sentence-level variant assigning a confidence to each claim. Both are trained using the GRPO algorithm with reward functions based on the logarithmic scoring rule, which incentivizes truthful self-assessment by penalizing miscalibration and guarantees optimal calibration under reward maximization. Experimentally, ConRad substantially improves calibration and outperforms competing methods. In a clinical evaluation we show that ConRad's report level scores are well aligned with clinicians' judgment. By highlighting full reports or low-confidence statements for targeted review, ConRad can support safer clinical integration of AI-assistance for report generation.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Medical vision-language models, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 24 of its 31 references that are in the map. First posted on arXiv 31 Mar 2026.

Builds on

24 of 31 references.

Closest ancestors · 1

Same direction, newest first

  1. 2023CXR-LLAVA: a multimodal large language model for interpreting chest X-ray images22 Oct

Other references · 23

  1. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  2. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  3. 2022Language Models (Mostly) Know What They Knowcited by 1,783
  4. 2023The Internal State of an LLM Knows When It's Lyingcited by 644
  5. 2019MIMIC-CXR-JPG, a large publicly available database of labeled chest radiographscited by 590
  6. 2025MedGemma Technical Reportcited by 398
  7. 2023Generating with Confidence: Uncertainty Quantification for Black-box Large Language Modelscited by 286
  8. 2023A Survey of Confidence Estimation and Calibration in Large Language Modelscited by 253
  9. 2024MAIRA-2: Grounded Radiology Report Generationcited by 140
  10. 2024GREEN: Generative Radiology Report Evaluation and Error Notationcited by 124
  11. 2023R-Tuning: Instructing Large Language Models to Say `I Don't Know'cited by 106
  12. 2024Taming Overconfidence in LLMs: Reward Calibration in RLHFcited by 105
11 more
  1. 2024Unfamiliar Finetuning Examples Control How Language Models Hallucinatecited by 85
  2. 2023Calibration in Deep Learning: A Survey of the State-of-the-Artcited by 72
  3. 2025Rewarding Doubt: A Reinforcement Learning Approach to Calibrated Confidence Expression of Large Language Modelscited by 48
  4. 2023Look Before You Leap: An Exploratory Study of Uncertainty Measurement for Large Language Modelscited by 40
  5. 2023Rad-ReStruct: A Novel VQA Benchmark and Method for Structured Radiology Reportingcited by 17
  6. 2025LoVeC: Reinforcement Learning for Better Verbalized Confidence in Long-Form Generationscited by 17
  7. 2021Trust It or Not: Confidence-Guided Automatic Radiology Report Generationcited by 9
  8. 2024Semantic Consistency-Based Uncertainty Quantification for Factuality in Radiology Report Generationcited by 4
  9. 2024The Impact of AI Assistance on Radiology Reporting: A Pilot Study Using Simulated AI Draft Reportscited by 4
  10. 2025Delusions of Large Language Modelscited by 4
  11. 2024Quality Control for Radiology Report Generation Models via Auxiliary Auditing Componentscited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025RadHiera: Semantic Hierarchical Reinforcement Learning for Medical Report Generation
  2. 2026Enhancing Reinforcement Learning for Radiology Report Generation with Evidence-aware Rewards and Self-correcting Preference Learning
  3. 2026Calibrated Triage, Not Autonomy: Confidence Estimation for Medical Vision-Language Models
  4. 2025Reason Like a Radiologist: Chain-of-Thought and Reinforcement Learning for Verifiable Report Generation
  5. 2025Medical AI Consensus: A Multi-Agent Framework for Radiology Report Generation and Evaluation
  6. 2026HERO: Hierarchical Evidential Reasoning Optimization for Radiology Report Generation via Reason-then-Summarize
  7. 2026UR2-MLLM: Uncertainty-aware Revisit Reasoning in Multimodal Large Language Models for Radiology Report Generation
  8. 2025Prompt4Trust: A Reinforcement Learning Prompt Augmentation Framework for Clinically-Aligned Confidence Calibration in Multimodal Large Language Models