aiqbench analysis
This research introduces PDD-RRG, a novel framework that adds a posterior diagnostic decision stage to radiology report generation models without requiring retraining. By leveraging Bayesian posterior probabilities and learned thresholds across multiple data subsets, the method effectively integrates conflicting diagnoses from different input perspectives to refine clinical conclusions. The approach demonstrates improved efficacy on MIMIC-CXR by mitigating errors associated with indiscriminately incorporating additional medical imaging or historical data.
PDD-RRG: Posterior Diagnostic Decision for Study-level Radiology Report Generation
arXiv:2608.03055v1 Announce Type: cross Abstract: Automatic radiology report generation (RRG) aims to simulate the workflow of radiologists, assisting them in clinical diagnosis. However, existing methods often fall short in utilizing all information relevant to the examination, as is typically done in clinical practice. Although some works attempt to incorporate multi-view images and historical data, these additional inputs may sometimes lead to avoidable diagnostic errors on the contrary. To address these challenges, we introduce a decision-making stage after report generation for the first time and propose a Posterior Diagnostic Decision framework (PDD-RRG) to integrate potentially conflicting diagnoses. Specifically, we create various subsets of input data and utilize an existing RRG model to generate reports from different perspectives. Then the Bayesian posterior probability and the learned thresholds for each clinical observation are calculated to obtain an aggregated diagnostic conclusion, which is subsequently used to refine the generated report. Experiments on MIMIC-CXR demonstrate that our proposed PDD-RRG can effectively enhance the clinical efficacy of existing RRG models without any retraining.
Abstract:Automatic radiology report generation (RRG) aims to simulate the workflow of radiologists, assisting them in clinical diagnosis. However, existing methods often fall short in utilizing all information relevant to the examination, as is typically done in clinical practice. Although some works attempt to incorporate multi-view images and historical data, these additional inputs may sometimes lead to avoidable diagnostic errors on the contrary. To address these challenges, we introduce a decision-making stage after report generation for the first time and propose a Posterior Diagnostic Decision framework (PDD-RRG) to integrate potentially conflicting diagnoses. Specifically, we create various subsets of input data and utilize an existing RRG model to generate reports from different perspectives. Then the Bayesian posterior probability and the learned thresholds for each clinical observation are calculated to obtain an aggregated diagnostic conclusion, which is subsequently used to refine the generated report. Experiments on MIMIC-CXR demonstrate that our proposed PDD-RRG can effectively enhance the clinical efficacy of existing RRG models without any retraining.
Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
arXiv:2608.03055 [cs.CV]
(or arXiv:2608.03055v1 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2608.03055 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 263 words
Related coverage
Introducing CARE-X: Towards Clinically Useful Radiology VLMs with Auxiliary Supervision, Reward-Aligned Learning, and Tool-Augmented Measurement
Microsoft Research · Aug 11, 2026
Introducing NV-Reason-CT Open 3D CT VLM for Radiologist Chain-of-Thought Reasoning
NVIDIA · Sep 23, 2026
MedPAIR: Measuring Whether Physicians and AI Agree on What Matters in Medical QA
Snorkel AI Blog · Oct 2, 2026
From Sampled Evidence to Radiology Reports: Objectives, Constraints, and Realization
arXiv · Oct 14, 2026