Localizing Before Answering: A Hallucination Evaluation Benchmark for Grounded Medical Multimodal LLMs

• URL: http://arxiv.org/abs/2505.00744v4
• Date: Tue, 15 Jul 2025 14:41:15 GMT
• Title: Localizing Before Answering: A Hallucination Evaluation Benchmark for Grounded Medical Multimodal LLMs
• Abstract summary: We introduce HEAL-MedVQA, a benchmark designed to evaluate LMMs' localization abilities and robustness.<n>We propose the Localize-before-Answer (LobA) framework, which trains LMMs to localize target regions of interest and self-prompt to emphasize segmented pathological areas.<n> Experimental results demonstrate that our approach significantly outperforms state-of-the-art biomedical LMMs on the challenging HEAL-MedVQA benchmark.
• Score: 18.43222146132849
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Medical Large Multi-modal Models (LMMs) have demonstrated remarkable capabilities in medical data interpretation. However, these models frequently generate hallucinations contradicting source evidence, particularly due to inadequate localization reasoning. This work reveals a critical limitation in current medical LMMs: instead of analyzing relevant pathological regions, they often rely on linguistic patterns or attend to irrelevant image areas when responding to disease-related queries. To address this, we introduce HEAL-MedVQA (Hallucination Evaluation via Localization MedVQA), a comprehensive benchmark designed to evaluate LMMs' localization abilities and hallucination robustness. HEAL-MedVQA features (i) two innovative evaluation protocols to assess visual and textual shortcut learning, and (ii) a dataset of 67K VQA pairs, with doctor-annotated anatomical segmentation masks for pathological regions. To improve visual reasoning, we propose the Localize-before-Answer (LobA) framework, which trains LMMs to localize target regions of interest and self-prompt to emphasize segmented pathological areas, generating grounded and reliable answers. Experimental results demonstrate that our approach significantly outperforms state-of-the-art biomedical LMMs on the challenging HEAL-MedVQA benchmark, advancing robustness in medical VQA.

Related papers

• Enhancing Medical Visual Grounding via Knowledge-guided Spatial Prompts [52.256130375429414]
  Medical Visual Grounding (MVG) aims to identify relevant phrases from free-text radiology reports and localize their corresponding regions in medical images.<n>We present KnowMVG, a knowledge-enhanced prompting strategy that encodes phrase related medical knowledge into compact embeddings.<n>This design bridges high-level semantic understanding and fine-grained visual perception without introducing extra textual reasoning overhead.
  arXiv  Detail & Related papers  (2026-04-02T11:31:30Z)
• MedAlign: A Synergistic Framework of Multimodal Preference Optimization and Federated Meta-Cognitive Reasoning [52.064286116035134]
  We develop MedAlign, a framework to ensure visually accurate LVLM responses for Medical Visual Question Answering (Med-VQA)<n>We first propose a multimodal Direct Preference Optimization (mDPO) objective to align preference learning with visual context.<n>We then design a Retrieval-Aware Mixture-of-Experts (RA-MoE) architecture that utilizes image and text similarity to route queries to a specialized and context-augmented LVLM.
  arXiv  Detail & Related papers  (2025-10-24T02:11:05Z)
• Proactive Reasoning-with-Retrieval Framework for Medical Multimodal Large Language Models [15.530083855947987]
  We propose the first Multimodal Medical Reasoning-with-Retrieval framework, Med-RwR.<n>Med-RwR actively retrieves external knowledge by querying observed symptoms or domain-specific medical concepts during reasoning.<n> Evaluation on various public medical benchmarks demonstrates Med-RwR's significant improvements over baseline models.
  arXiv  Detail & Related papers  (2025-10-21T05:18:18Z)
• TemMed-Bench: Evaluating Temporal Medical Image Reasoning in Vision-Language Models [54.48710348910535]
  Existing medical reasoning benchmarks primarily focus on analyzing a patient's condition based on an image from a single visit.<n>We introduce TemMed-Bench, the first benchmark designed for analyzing changes in patients' conditions between different clinical visits.
  arXiv  Detail & Related papers  (2025-09-29T17:51:26Z)
• Lingshu: A Generalist Foundation Model for Unified Multimodal Medical Understanding and Reasoning [57.873833577058]
  We build a multimodal dataset enriched with extensive medical knowledge.<n>We then introduce our medical-specialized MLLM: Lingshu.<n>Lingshu undergoes multi-stage training to embed medical expertise and enhance its task-solving capabilities.
  arXiv  Detail & Related papers  (2025-06-08T08:47:30Z)
• Structured Outputs Enable General-Purpose LLMs to be Medical Experts [50.02627258858336]
  Large language models (LLMs) often struggle with open-ended medical questions.<n>We propose a novel approach utilizing structured medical reasoning.<n>Our approach achieves the highest Factuality Score of 85.8, surpassing fine-tuned models.
  arXiv  Detail & Related papers  (2025-03-05T05:24:55Z)
• MedHEval: Benchmarking Hallucinations and Mitigation Strategies in Medical Large Vision-Language Models [37.78272983522441]
  Large Vision Language Models (LVLMs) are becoming increasingly important in the medical domain.<n>MedHEval is a novel benchmark that systematically evaluates hallucinations and mitigation strategies in Med-LVLMs.<n>We conduct experiments across 11 popular (Med)-LVLMs and evaluate 7 state-of-the-art hallucination mitigation techniques.
  arXiv  Detail & Related papers  (2025-03-04T00:40:09Z)
• Fact or Guesswork? Evaluating Large Language Models' Medical Knowledge with Structured One-Hop Judgments [108.55277188617035]
  Large language models (LLMs) have been widely adopted in various downstream task domains, but their abilities to directly recall and apply factual medical knowledge remains under-explored.<n>We introduce the Medical Knowledge Judgment dataset (MKJ), a dataset derived from the Unified Medical Language System (UMLS), a comprehensive repository of standardized vocabularies and knowledge graphs.<n>Through a binary classification framework, MKJ evaluates LLMs' grasp of fundamental medical facts by having them assess the validity of concise, one-hop statements.
  arXiv  Detail & Related papers  (2025-02-20T05:27:51Z)
• Improving Medical Large Vision-Language Models with Abnormal-Aware Feedback [57.98393950821579]
  We propose a novel UMed-LVLM designed to unveil medical abnormalities.<n>We propose a prompt method utilizing the GPT-4V to generate diagnoses based on identified abnormal areas in medical images.<n>Our UMed-LVLM significantly outperforms existing Med-LVLMs in identifying and understanding medical abnormalities.
  arXiv  Detail & Related papers  (2025-01-02T17:37:20Z)
• LLM-MedQA: Enhancing Medical Question Answering through Case Studies in Large Language Models [18.6994780408699]
  Large Language Models (LLMs) face significant challenges in medical question answering.<n>We propose a novel approach incorporating similar case generation within a multi-agent medical question-answering system.<n>Our method capitalizes on the model's inherent medical knowledge and reasoning capabilities, eliminating the need for additional training data.
  arXiv  Detail & Related papers  (2024-12-31T19:55:45Z)
• MedHallBench: A New Benchmark for Assessing Hallucination in Medical Large Language Models [0.0]
  Medical Large Language Models (MLLMs) have demonstrated potential in healthcare applications.<n>Their propensity for hallucinations presents substantial risks to patient care.<n>This paper introduces MedHallBench, a comprehensive benchmark framework for evaluating and mitigating hallucinations in MLLMs.
  arXiv  Detail & Related papers  (2024-12-25T16:51:29Z)
• Comprehensive and Practical Evaluation of Retrieval-Augmented Generation Systems for Medical Question Answering [70.44269982045415]
  Retrieval-augmented generation (RAG) has emerged as a promising approach to enhance the performance of large language models (LLMs) We introduce Medical Retrieval-Augmented Generation Benchmark (MedRGB) that provides various supplementary elements to four medical QA datasets. Our experimental results reveals current models' limited ability to handle noise and misinformation in the retrieved documents.
  arXiv  Detail & Related papers  (2024-11-14T06:19:18Z)
• Detecting and Evaluating Medical Hallucinations in Large Vision Language Models [22.30139330566514]
  Large Vision Language Models (LVLMs) are increasingly integral to healthcare applications. LVLMs inherit susceptibility to hallucinations-a significant concern in high-stakes medical contexts. We introduce Med-HallMark, the first benchmark specifically designed for hallucination detection and evaluation. We also present MediHallDetector, a novel Medical LVLM engineered for precise hallucination detection.
  arXiv  Detail & Related papers  (2024-06-14T17:14:22Z)
• MedExQA: Medical Question Answering Benchmark with Multiple Explanations [2.2246416434538308]
  This paper introduces MedExQA, a novel benchmark in medical question-answering to evaluate large language models' (LLMs) understanding of medical knowledge through explanations. By constructing datasets across five distinct medical specialties, we address a major gap in current medical QA benchmarks. Our work highlights the importance of explainability in medical LLMs, proposes an effective methodology for evaluating models beyond classification accuracy, and sheds light on one specific domain, speech language pathology.
  arXiv  Detail & Related papers  (2024-06-10T14:47:04Z)
• Worse than Random? An Embarrassingly Simple Probing Evaluation of Large Multimodal Models in Medical VQA [24.10436440624249]
  Large Multimodal Models (LMMs) have shown remarkable progress in medical Visual Question Answering (Med-VQA) This study reveals that when subjected to simple probing evaluation, state-of-the-art models perform worse than random guessing on medical diagnosis questions.
  arXiv  Detail & Related papers  (2024-05-30T18:56:01Z)
• RJUA-MedDQA: A Multimodal Benchmark for Medical Document Question Answering and Clinical Reasoning [14.366349078707263]
  RJUA-MedDQA is a comprehensive benchmark in the field of medical specialization. This work introduces RJUA-MedDQA, a comprehensive benchmark in the field of medical specialization.
  arXiv  Detail & Related papers  (2024-02-19T06:57:02Z)
• AI Hospital: Benchmarking Large Language Models in a Multi-agent Medical Interaction Simulator [69.51568871044454]
  We introduce textbfAI Hospital, a framework simulating dynamic medical interactions between emphDoctor as player and NPCs. This setup allows for realistic assessments of LLMs in clinical scenarios. We develop the Multi-View Medical Evaluation benchmark, utilizing high-quality Chinese medical records and NPCs.
  arXiv  Detail & Related papers  (2024-02-15T06:46:48Z)
• OmniMedVQA: A New Large-Scale Comprehensive Evaluation Benchmark for Medical LVLM [48.16696073640864]
  We introduce OmniMedVQA, a novel comprehensive medical Visual Question Answering (VQA) benchmark. All images in this benchmark are sourced from authentic medical scenarios. We have found that existing LVLMs struggle to address these medical VQA problems effectively.
  arXiv  Detail & Related papers  (2024-02-14T13:51:56Z)
• Cross-Modal Causal Intervention for Medical Report Generation [109.83549148448469]
  Medical report generation (MRG) is essential for computer-aided diagnosis and medication guidance. Due to the spurious correlations within image-text data induced by visual and linguistic biases, it is challenging to generate accurate reports reliably describing lesion areas. We propose a novel Visual-Linguistic Causal Intervention (VLCI) framework for MRG, which consists of a visual deconfounding module (VDM) and a linguistic deconfounding module (LDM)
  arXiv  Detail & Related papers  (2023-03-16T07:23:55Z)

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.