ProMedical: Hierarchical Fine-Grained Criteria Modeling for Medical LLM Alignment via Explicit Injection

• URL: http://arxiv.org/abs/2604.08326v1
• Date: Thu, 09 Apr 2026 14:57:33 GMT
• Title: ProMedical: Hierarchical Fine-Grained Criteria Modeling for Medical LLM Alignment via Explicit Injection
• Abstract summary: We introduce ProMedical, a unified alignment framework grounded in fine-grained clinical criteria.<n>We first construct ProMedical-Preference-50k, a dataset generated via a human-in-the-loop pipeline.<n>Unlike traditional scalar reward models, our approach explicitly disentangles safety constraints from general proficiency.
• Score: 1.1273389819207937
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Aligning Large Language Models (LLMs) with high-stakes medical standards remains a significant challenge, primarily due to the dissonance between coarse-grained preference signals and the complex, multi-dimensional nature of clinical protocols. To bridge this gap, we introduce ProMedical, a unified alignment framework grounded in fine-grained clinical criteria. We first construct ProMedical-Preference-50k, a dataset generated via a human-in-the-loop pipeline that augments medical instructions with rigorous, physician-derived rubrics. Leveraging this corpus, we propose the Explicit Criteria Injection paradigm to train a multi-dimensional reward model. Unlike traditional scalar reward models, our approach explicitly disentangles safety constraints from general proficiency, enabling precise guidance during reinforcement learning. To rigorously validate this framework, we establish ProMedical-Bench, a held-out evaluation suite anchored by double-blind expert adjudication. Empirical evaluations demonstrate that optimizing the Qwen3-8B base model via ProMedical-RM-guided GRPO yields substantial gains, improving overall accuracy by 22.3% and safety compliance by 21.7%, effectively rivaling proprietary frontier models. Furthermore, the aligned policy generalizes robustly to external benchmarks, demonstrating performance comparable to state-of-the-art models on UltraMedical. We publicly release our datasets, reward models, and benchmarks to facilitate reproducible research in safety-aware medical alignment.

Related papers

• Concept-Grounded Reasoning with Prompt-Driven Localization for Interpretable Structured Report Generation [66.62662330558906]
  CORAL is a multimodal framework that integrates spatial grounding and concept-level supervision into a unified reasoning process.<n>CORAL employs a prompt-driven medical segmentation model to localize lesions and predicts multi-class clinical attributes through a Concept Bottleneck module.<n>Experiments on BUS-CoT and IU X-ray datasets demonstrate consistent improvements in diagnostic accuracy, concept consistency, and report quality over strong general-purpose and medical MLLMs.
  arXiv  Detail & Related papers  (2026-09-14T10:21:23Z)
• DobicVLM: Aligning Chest X-Ray Report Generation with Clinically-Grounded Programmatic Rewards via Group Relative Policy Optimization [0.4418222884390162]
  We introduce DobicVLM, a vision-language model combining supervised fine-tuning on MedGemma-4B with Group Relative Policy Optimization.<n>Our approach uses interpretable, rule-based reward components; structural verification, anatomical checklist, semantic similarity, and length constraints.<n>DobicVLM outperforms Gemini 2.5 Flash across the majority of criteria, achieving the highest impression accuracy (27.2%) and medical terminology (86.5%) compared to both Gemini 2.5 Flash and MedGemma 4B baselines.
  arXiv  Detail & Related papers  (2026-07-21T11:20:09Z)
• Latent-CURE for Breast Cancer Diagnosis [16.345145740430457]
  We propose Latent-CURE, a novel diagnostic framework driven by asymmetric weighted chain-of-thought methodology.<n>Unlike traditional approaches, our framework constructs an implicit reasoning trajectory forcing the model to sequentially infer standardized BI-RADS morphological descriptors.
  arXiv  Detail & Related papers  (2026-06-29T08:05:16Z)
• Analyzing and Improving Fine-grained Preference Optimization in Medical LVLMs [53.697403481143404]
  Large Vision-Language Models (LVLMs) have achieved strong performance across medical imaging tasks.<n>But they remain prone to factual inconsistencies, poor visual grounding, and misalignment with clinically meaningful feedback.
  arXiv  Detail & Related papers  (2026-06-10T18:35:36Z)
• Clinical Cognition Alignment for Gastrointestinal Diagnosis with Multimodal LLMs [63.535652574541764]
  Multimodal Large Language Models (MLLMs) have demonstrated remarkable potential in medical image analysis.<n>Their application in gastrointestinal endoscopy is currently hindered by two critical limitations.<n>We propose a novel Clinical-Cognitive-Aligned (CogAlign) framework to address these challenges.
  arXiv  Detail & Related papers  (2026-03-21T07:47:37Z)
• ClinAlign: Scaling Healthcare Alignment from Clinician Preference [46.61989657665566]
  We introduce HealthRubrics, a dataset of 7,034 physician-verified preference examples.<n>We distill these rubrics into HealthPrinciples: 119 broadly reusable, clinically grounded principles.<n>A 30B-A3B model trained with our framework achieves 33.4% on HealthBench-Hard.
  arXiv  Detail & Related papers  (2026-02-10T11:02:57Z)
• MMedExpert-R1: Strengthening Multimodal Medical Reasoning via Domain-Specific Adaptation and Clinical Guideline Reinforcement [63.82954136824963]
  Medical Vision-Language Models excel at perception tasks with complex clinical reasoning required in real-world scenarios.<n>We propose a novel reasoning MedVLM that addresses these challenges through domain-specific adaptation and guideline reinforcement.
  arXiv  Detail & Related papers  (2026-01-16T02:32:07Z)
• Multidimensional Rubric-oriented Reward Model Learning via Geometric Projection Reference Constraints [4.79357178898034]
  We introduce MR-RML (Multidimensional-oriented Reward Model Learning) with GPRC (Geometric Projection Reference Constraints)<n>Our approach introduces three key innovations: (1) a medical standard system that embeds domain-specific guidelines throughout the training pipeline; (2) an independent multi-dimensional reward model that decomposes evaluation criteria; and (3) projection reference constraints that translate clinical cognitive logic into mathematical regularization.<n>Our method significantly boosts the performance of the base Qwen-32B model, with improvements of 45% on the full subset and 85% on the hard subset.
  arXiv  Detail & Related papers  (2025-11-20T08:26:16Z)
• Baichuan-M2: Scaling Medical Capability with Large Verifier System [40.86227022086866]
  We introduce a novel dynamic verification framework that moves beyond static answer verifier.<n>We develop Baichuan-M2, a medical augmented reasoning model trained through a multi-stage reinforcement learning strategy.<n> evaluated on HealthBench, Baichuan-M2 outperforms all other open-source models and most advanced closed-source counterparts.
  arXiv  Detail & Related papers  (2025-09-02T11:23:35Z)
• Uncertainty-Driven Expert Control: Enhancing the Reliability of Medical Vision-Language Models [52.2001050216955]
  Existing methods aim to enhance the performance of Medical Vision Language Model (MedVLM) by adjusting model structure, fine-tuning with high-quality data, or through preference fine-tuning.<n>We propose an expert-in-the-loop framework named Expert-Controlled-Free Guidance (Expert-CFG) to align MedVLM with clinical expertise without additional training.
  arXiv  Detail & Related papers  (2025-07-12T09:03:30Z)
• Refine Medical Diagnosis Using Generation Augmented Retrieval and Clinical Practice Guidelines [16.56254046507092]
  We introduce GARMLE-G, a Generation-Augmented Retrieval framework that grounds medical language model outputs in authoritative guidelines.<n>Unlike conventional Retrieval-Augmented Generation based approaches, GARMLE-G enables hallucination-free outputs by directly retrieving authoritative guideline content.<n>A prototype system for hypertension diagnosis was developed and evaluated on multiple metrics, demonstrating superior retrieval precision, semantic relevance, and clinical guideline adherence.
  arXiv  Detail & Related papers  (2025-06-22T11:31:13Z)
• Towards a clinically accessible radiology foundation model: open-access and lightweight, with automated evaluation [113.5002649181103]
  Training open-source small multimodal models (SMMs) to bridge competency gaps for unmet clinical needs in radiology. For training, we assemble a large dataset of over 697 thousand radiology image-text pairs. For evaluation, we propose CheXprompt, a GPT-4-based metric for factuality evaluation, and demonstrate its parity with expert evaluation. The inference of LlaVA-Rad is fast and can be performed on a single V100 GPU in private settings, offering a promising state-of-the-art tool for real-world clinical applications.
  arXiv  Detail & Related papers  (2024-03-12T18:12:02Z)

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.