mlsanityData to 14 Oct 2026

Beyond Fluency: A Clinical Benchmark and Anomaly-Enhanced Baseline for Spine MRI Report Generation

arXiv 2608.07117 ↗First posted 7 Aug 2026cs.CVcited by 0

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

Abstract (arXiv)

Radiology reporting is time-consuming and subject to inter-rater variability, making automated report generation an attractive clinical application for Vision-Language Models (VLMs). We benchmark state-of-the-art VLMs on lumbar spine MRI with a focus on diagnostic accuracy and demonstrate that standard lexical and semantic metrics poorly reflect clinical correctness: fluent, well-structured reports can score highly while containing clinically meaningful diagnostic errors. To address this failure mode, we propose an architecture-agnostic framework that augments VLM inputs with spatially localized, disc-level anomaly heatmaps generated by a semi-supervised U-Net++ model. These heatmaps both improve anatomical sensitivity through explicit visual grounding and provide an independent interpretability output for clinical oversight, moving us closer to diagnostically reliable, visually grounded VLMs for lumbar spine MRI interpretation.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Medical vision-language models, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 15 of its 39 references that are in the map. First posted on arXiv 7 Aug 2026.

Builds on

15 of 39 references.

Closest ancestors · 2

Same direction, newest first

  1. 2023LLaVA-Med: Training a Large Language-and-Vision Assistant for Biomedicine in One Day1 Jun
  2. 2023BiomedGPT: A Generalist Vision-Language Foundation Model for Diverse Biomedical Tasks26 May

Other references · 13

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2019BERTScore: Evaluating Text Generation with BERTcited by 5,512
  3. 2018UNet++: A Nested U-Net Architecture for Medical Image Segmentationcited by 1,525
  4. 2022Cross-modal Memory Networks for Radiology Report Generationcited by 185
  5. 2023R2GenGPT: Radiology Report Generation with Frozen LLMscited by 114
  6. 2023METransformer: Radiology Report Generation by Transformer with Multiple Learnable Expert Tokenscited by 101
  7. 2024CT2Rep: Automated Radiology Report Generation for 3D Medical Imagingcited by 96
  8. 2022Improving Chest X-Ray Report Generation by Leveraging Warm Startingcited by 95
  9. 2023LLM-CXR: Instruction-Finetuned LLM for CXR Image Understanding and Generationcited by 57
  10. 2024Automated Radiology Report Generation: A Review of Recent Advancescited by 43
  11. 2023Lumbar spine segmentation in MR images: a dataset and a public benchmarkcited by 30
  12. 2022SpineNetV2: Automated Detection, Labelling and Radiological Grading Of Clinical MR Scanscited by 7
1 more
  1. 2025The RSNA Lumbar Degenerative Imaging Spine Classification (LumbarDISC) Datasetcited by 6

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025SpineBench: Benchmarking Multimodal LLMs for Spinal Pathology Analysis
  2. 2025SpineBench: A Clinically Salient, Level-Aware Benchmark Powered by the SpineMed-450k Corpus
  3. 2026An Explainable Vision-Language Model Framework with Adaptive PID-Tversky Loss for Lumbar Spinal Stenosis Diagnosis
  4. 2026Toward Guarantees for Clinical Reasoning in Vision Language Models via Formal Verification
  5. 2026Evaluating the Diagnostic Robustness of Vision-Language Models Under Visual and Textual Perturbations
  6. 2026A multi-agent system for spine MRI report generation from multi-sequence imaging
  7. 2026RadSight: Towards Perceptually Reliable Multimodal Radiology Image Understanding
  8. 2026Generating Reports or Repeating Templates? Measuring and Mitigating Template Collapse in 3D CT Report Generation