MedGPT-oss: Training a General-Purpose Vision-Language Model for Biomedicine

• URL: http://arxiv.org/abs/2603.00842v1
• Date: Sun, 01 Mar 2026 00:06:43 GMT
• Title: MedGPT-oss: Training a General-Purpose Vision-Language Model for Biomedicine
• Abstract summary: We introduce MEDGPT-OSS, an open-weight, 20B- parameter vision-language model designed to facilitate open research in clinical AI.<n>Rather than relying on architectural complexity, MEDGPT-OSS pairs the GPT-oss language backbone with a visual front-end via a optimized, three-stage training curriculum.<n>It successfully outperforms larger open medical models on out-of-distribution multimodal reasoning and complex text-only clinical tasks.
• Score: 38.06252990946545
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Biomedical multimodal assistants have the potential to unify radiology, pathology, and clinical-text reasoning, yet a critical deployment gap remains: top-performing systems are either closed-source or computationally prohibitive, precluding the on-premises deployment required for patient privacy and PHI compliance. We introduce MEDGPT-OSS, an open-weight, 20B-parameter generalist vision-language model designed to facilitate open research in clinical AI. Rather than relying on architectural complexity, MEDGPT-OSS pairs the GPT-oss language backbone with a visual front-end via a optimized, three-stage training curriculum. By progressively domain-adapting these modules through rigorous data curation and long-context multimodal alignment, we demonstrate that a 20B model can bridge the capacity gap. It successfully outperforms larger open medical models on out-of-distribution (OOD) multimodal reasoning and complex text-only clinical tasks. By unifying diverse modalities under a single instruction-following interface, MEDGPT-OSS maintains a parameter-efficient footprint fully compatible with commodity GPUs. We release the complete training recipe, open-weight checkpoints, and a rigorous evaluation harness to serve as a verifiable foundation for privacy-preserving, institution-specific clinical AI research.

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)
• MedLayBench-V: A Large-Scale Benchmark for Expert-Lay Semantic Alignment in Medical Vision Language Models [2.0221473384338884]
  We introduce MedLayBench-V, the first large-scale multimodal benchmark dedicated to expert-lay semantic alignment.<n>MedLayBench-V provides a verified foundation for training and evaluating next-generation Med-VLMs capable of bridging the communication divide between clinical experts and patients.
  arXiv  Detail & Related papers  (2026-04-07T11:39:41Z)
• 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)
• Vision Foundry: A System for Training Foundational Vision AI Models [0.0]
  Vision Foundry is a code-free, HIPAA-compliant platform that democratizes pre-training, adaptation, and deployment of vision models.<n>By bridging the gap between advanced representation learning and practical application, Vision Foundry enables domain experts to develop state-of-the-art clinical AI tools.
  arXiv  Detail & Related papers  (2025-12-03T14:02:22Z)
• DuPLUS: Dual-Prompt Vision-Language Framework for Universal Medical Image Segmentation and Prognosis [5.494301428436596]
  We introduce DuPLUS, a deep learning framework for efficient multi-modal medical image analysis.<n>DuPLUS introduces a novel vision-language framework that leverages hierarchical semantic prompts for fine-grained control over the analysis task.<n>For segmentation, DuPLUS is able to generalize across three imaging modalities, ten different various medical datasets, encompassing more than 30 organs and tumor types.
  arXiv  Detail & Related papers  (2025-10-03T20:01:00Z)
• A Fully Open and Generalizable Foundation Model for Ultrasound Clinical Applications [77.3888788549565]
  We present EchoCare, a novel ultrasound foundation model for generalist clinical use.<n>We developed EchoCare via self-supervised learning on our curated, publicly available, large-scale dataset EchoCareData.<n>With minimal training, EchoCare outperforms state-of-the-art comparison models across 10 representative ultrasound benchmarks.
  arXiv  Detail & Related papers  (2025-09-15T10:05:31Z)
• An Agentic Model Context Protocol Framework for Medical Concept Standardization [5.12407270785129]
  We develop a zero-training, hallucination-preventive mapping system based on the Model Context Protocol (MCP)<n>The system enables explainable mapping and significantly improves efficiency and accuracy with minimal effort.
  arXiv  Detail & Related papers  (2025-09-04T02:32:22Z)
• Advancing AI Research Assistants with Expert-Involved Learning [84.30323604785646]
  Large language models (LLMs) and large multimodal models (LMMs) promise to accelerate biomedical discovery, yet their reliability remains unclear.<n>We introduce ARIEL (AI Research Assistant for Expert-in-the-Loop Learning), an open-source evaluation and optimization framework.<n>We find that state-of-the-art models generate fluent but incomplete summaries, whereas LMMs struggle with detailed visual reasoning.
  arXiv  Detail & Related papers  (2025-05-03T14:21:48Z)
• From large language models to multimodal AI: A scoping review on the potential of generative AI in medicine [40.23383597339471]
  multimodal AI is capable of integrating diverse data modalities, including imaging, text, and structured data, within a single model.<n>This scoping review explores the evolution of multimodal AI, highlighting its methods, applications, datasets, and evaluation in clinical settings.<n>Our findings underscore a shift from unimodal to multimodal approaches, driving innovations in diagnostic support, medical report generation, drug discovery, and conversational AI.
  arXiv  Detail & Related papers  (2025-02-13T11:57:51Z)
• Continually Evolved Multimodal Foundation Models for Cancer Prognosis [50.43145292874533]
  Cancer prognosis is a critical task that involves predicting patient outcomes and survival rates.<n>Previous studies have integrated diverse data modalities, such as clinical notes, medical images, and genomic data, leveraging their complementary information.<n>Existing approaches face two major limitations. First, they struggle to incorporate newly arrived data with varying distributions into training, such as patient records from different hospitals.<n>Second, most multimodal integration methods rely on simplistic concatenation or task-specific pipelines, which fail to capture the complex interdependencies across modalities.
  arXiv  Detail & Related papers  (2025-01-30T06:49:57Z)
• 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.