Search papers, labs, and topics across Lattice.
This study introduces DRRG, a discrete diffusion framework for automatic radiology report generation that allows for iterative refinement of reports, addressing the limitations of traditional autoregressive methods. By utilizing a clinical-entities-aware complementary mask and a concept-conditioning module, DRRG enhances token supervision and integrates image-derived clinical concepts into its generation process. The results show that DRRG outperforms existing methods on multiple metrics across two datasets, demonstrating its potential for improving report quality and consistency in radiology.
Iterative refinement in radiology report generation can significantly enhance accuracy, with DRRG outperforming traditional methods despite using a smaller model.
Purpose: Automatic radiology report generation (RRG) has been widely explored to improve reporting accuracy and reduce radiologists'workload. Most existing methods rely on autoregressive (AR) frameworks that generate reports token by token and cannot revise earlier content, making them prone to error propagation and inconsistent with the iterative refinement process of radiological reporting. In contrast, discrete diffusion large language models (DLLMs) generate text through iterative denoising, naturally enabling report refinement. However, DLLMs have not been extensively investigated for RRG. In this study, we developed and evaluated a discrete diffusion framework for RRG that enables iterative refinement rather than conventional left-to-right autoregressive decoding. Materials and methods: We developed DRRG, a DLLM-based framework that formulates RRG as iterative masked-token denoising. DRRG incorporates a clinical-entities-aware complementary mask to improve token supervision coverage and emphasize clinically important entities, together with a concept-conditioning module that injects image-derived clinical concepts into visual representations. DRRG was trained and evaluated on MIMIC-CXR and CheXpert Plus. Results: On MIMIC-CXR, DRRG achieved BLEU-4 of 0.210, CheXpert-F1 of 0.549, RadGraph-F1 of 0.281, GREEN of 0.360, and RaTEScore of 0.604, outperforming the compared methods on most reported metrics, despite employing a substantially smaller LLM decoder. On CheXpert Plus, DRRG achieved the highest BLEU-4 (0.119) and CheXpert-F1 (0.347) among the compared methods. Conclusion: Discrete diffusion provides an effective alternative to autoregressive radiology report generation by enabling iterative, bidirectional report refinement. Incorporating clinically focused masking and image-derived concept conditioning improves report quality and clinical consistency.
