$μ^2$Tokenizer: Differentiable Multi-Scale Multi-Modal Tokenizer for Radiology Report Generation

• URL: http://arxiv.org/abs/2507.00316v2
• Date: Wed, 02 Jul 2025 01:08:41 GMT
• Title: $μ^2$Tokenizer: Differentiable Multi-Scale Multi-Modal Tokenizer for Radiology Report Generation
• Authors: Siyou Li, Pengyao Qin, Huanan Wu, Dong Nie, Arun J. Thirunavukarasu, Juntao Yu, Le Zhang,
• Abstract summary: We propose $mu2$LLM, a $underlinetextbfmu$ltiscale $underlinetextbfmu$ltimodal large language models for radiology report generation tasks.<n>The novel $mu2$Tokenizer, as an intermediate layer, integrates multi-modal features from the multiscale visual tokenizer and the text tokenizer.<n>For prompt engineering, we introduce a five-stage, LLM-driven pipeline that converts routine CT reports into paired visual-question-answer triples and citation-linked reasoning narratives.
• Score: 9.947108972979155
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Automated radiology report generation (RRG) aims to produce detailed textual reports from clinical imaging, such as computed tomography (CT) scans, to improve the accuracy and efficiency of diagnosis and provision of management advice. RRG is complicated by two key challenges: (1) inherent complexity in extracting relevant information from imaging data under resource constraints, and (2) difficulty in objectively evaluating discrepancies between model-generated and expert-written reports. To address these challenges, we propose $\mu^2$LLM, a $\underline{\textbf{mu}}$ltiscale $\underline{\textbf{mu}}$ltimodal large language models for RRG tasks. The novel ${\mu}^2$Tokenizer, as an intermediate layer, integrates multi-modal features from the multiscale visual tokenizer and the text tokenizer, then enhances report generation quality through direct preference optimization (DPO), guided by GREEN-RedLlama. Experimental results on four large CT image-report medical datasets demonstrate that our method outperforms existing approaches, highlighting the potential of our fine-tuned $\mu^2$LLMs on limited data for RRG tasks. At the same time, for prompt engineering, we introduce a five-stage, LLM-driven pipeline that converts routine CT reports into paired visual-question-answer triples and citation-linked reasoning narratives, creating a scalable, high-quality supervisory corpus for explainable multimodal radiology LLM. All code, datasets, and models will be publicly available in our official repository. https://github.com/Siyou-Li/u2Tokenizer

Related papers

• A Multimodal Multi-Agent Framework for Radiology Report Generation [2.1477122604204433]
  Radiology report generation (RRG) aims to automatically produce diagnostic reports from medical images.<n>We propose a multimodal multi-agent framework for RRG that aligns with the stepwise clinical reasoning workflow.
  arXiv  Detail & Related papers  (2025-05-14T20:28:04Z)
• Leveraging LLMs for Multimodal Retrieval-Augmented Radiology Report Generation via Key Phrase Extraction [0.0]
  We propose a retrieval-augmented generation approach that leverages multimodal retrieval and large language models.<n>Our method uses LLMs to extract key phrases from radiology reports, effectively focusing on essential diagnostic information.<n>We evaluate our approach on MIMIC-CXR dataset, achieving state-of-the-art results on CheXbert metrics and competitive RadGraph F1 metric.
  arXiv  Detail & Related papers  (2025-04-10T03:14:01Z)
• MM-OR: A Large Multimodal Operating Room Dataset for Semantic Understanding of High-Intensity Surgical Environments [49.45034796115852]
  Operating rooms (ORs) are complex, high-stakes environments requiring precise understanding of interactions among medical staff, tools, and equipment.<n>Current datasets fall short in scale, realism and do not capture the nature of OR scenes, limiting multimodal in OR modeling.<n>We introduce MM-OR, a realistic and large-scale multimodal OR dataset, and first dataset to enable multimodal scene graph generation.
  arXiv  Detail & Related papers  (2025-03-04T13:00:52Z)
• MINIMA: Modality Invariant Image Matching [52.505282811925454]
  We present MINIMA, a unified image matching framework for multiple cross-modal cases.<n>We scale up the modalities from cheap but rich RGB-only matching data, by means of generative models.<n>With MD-syn, we can directly train any advanced matching pipeline on randomly selected modality pairs to obtain cross-modal ability.
  arXiv  Detail & Related papers  (2024-12-27T02:39:50Z)
• LLM-RG4: Flexible and Factual Radiology Report Generation across Diverse Input Contexts [14.72366043711941]
  Current radiology report generation models are constrained to a fixed task paradigm.<n>We propose a novel large language model (LLM) based RRG framework, namely LLM-RG4.<n>We show that our model has minimal input-agnostic hallucinations, whereas current open-source models commonly suffer from this problem.
  arXiv  Detail & Related papers  (2024-12-16T17:29:51Z)
• Semantic Consistency-Based Uncertainty Quantification for Factuality in Radiology Report Generation [20.173287130474797]
  generative medical Vision Large Language Models (VLLMs) are prone to hallucinations and can produce inaccurate diagnostic information.<n>We introduce a novel Semantic Consistency-Based Uncertainty Quantification framework that provides both report-level and sentence-level uncertainties.<n>Our approach improves factuality scores by $10$%, achieved by rejecting $20$% of reports using the textttRadialog model on the MIMIC-CXR dataset.
  arXiv  Detail & Related papers  (2024-12-05T20:43:39Z)
• Large Language Models for Multimodal Deformable Image Registration [50.91473745610945]
  We propose a novel coarse-to-fine MDIR framework,LLM-Morph, for aligning the deep features from different modal medical images. Specifically, we first utilize a CNN encoder to extract deep visual features from cross-modal image pairs, then we use the first adapter to adjust these tokens, and use LoRA in pre-trained LLMs to fine-tune their weights. Third, for the alignment of tokens, we utilize other four adapters to transform the LLM-encoded tokens into multi-scale visual features, generating multi-scale deformation fields and facilitating the coarse-to-fine MDIR task
  arXiv  Detail & Related papers  (2024-08-20T09:58:30Z)
• R2GenCSR: Retrieving Context Samples for Large Language Model based X-ray Medical Report Generation [7.4871243017824165]
  This paper proposes a novel context-guided efficient X-ray medical report generation framework. Specifically, we introduce the Mamba as the vision backbone with linear complexity, and the performance obtained is comparable to that of the strong Transformer model.
  arXiv  Detail & Related papers  (2024-08-19T07:15:11Z)
• Unify, Align and Refine: Multi-Level Semantic Alignment for Radiology Report Generation [48.723504098917324]
  We propose an Unify, Align and then Refine (UAR) approach to learn multi-level cross-modal alignments. We introduce three novel modules: Latent Space Unifier, Cross-modal Representation Aligner and Text-to-Image Refiner. Experiments and analyses on IU-Xray and MIMIC-CXR benchmark datasets demonstrate the superiority of our UAR against varied state-of-the-art methods.
  arXiv  Detail & Related papers  (2023-03-28T12:42:12Z)
• Cross-Modal Causal Intervention for Medical Report Generation [107.76649943399168]
  Radiology Report Generation (RRG) is essential for computer-aided diagnosis and medication guidance.<n> generating accurate lesion descriptions remains challenging due to spurious correlations from visual-linguistic biases.<n>We propose a two-stage framework named CrossModal Causal Representation Learning (CMCRL)<n> Experiments on IU-Xray and MIMIC-CXR show that our CMCRL pipeline significantly outperforms state-of-the-art methods.
  arXiv  Detail & Related papers  (2023-03-16T07:23:55Z)
• MuRAG: Multimodal Retrieval-Augmented Generator for Open Question Answering over Images and Text [58.655375327681774]
  We propose the first Multimodal Retrieval-Augmented Transformer (MuRAG) MuRAG accesses an external non-parametric multimodal memory to augment language generation. Our results show that MuRAG achieves state-of-the-art accuracy, outperforming existing models by 10-20% absolute on both datasets.
  arXiv  Detail & Related papers  (2022-10-06T13:58:03Z)
• Brain Image Synthesis with Unsupervised Multivariate Canonical CSC$\ell_4$Net [122.8907826672382]
  We propose to learn dedicated features that cross both intre- and intra-modal variations using a novel CSC$ell_4$Net.
  arXiv  Detail & Related papers  (2021-03-22T05:19:40Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.