Paper
MedQ-Engine: A Closed-Loop Data Engine for Evolving MLLMs in Medical Image Quality Assessment
Authors
Jiyao Liu, Junzhi Ning, Wanying Qu, Lihao Liu, Chenglong Ma, Junjun He, Ningsheng Xu
Abstract
Medical image quality assessment (Med-IQA) is a prerequisite for clinical AI deployment, yet multimodal large language models (MLLMs) still fall substantially short of human experts, particularly when required to provide descriptive assessments with clinical reasoning beyond simple quality scores. However, improving them is hindered by the high cost of acquiring descriptive annotations and by the inability of one-time data collection to adapt to the model's evolving weaknesses. To address these challenges, we propose MedQ-Engine, a closed-loop data engine that iteratively evaluates the model to discover failure prototypes via data-driven clustering, explores a million-scale image pool using these prototypes as retrieval anchors with progressive human-in-the-loop annotation, and evolves through quality-assured fine-tuning, forming a self-improving cycle. Models are evaluated on complementary perception and description tasks. An entropy-guided routing mechanism triages annotations to minimize labeling cost. Experiments across five medical imaging modalities show that MedQ-Engine elevates an 8B-parameter model to surpass GPT-4o by over 13% and narrow the gap with human experts to only 4.34%, using only 10K annotations with more than 4x sample efficiency over random sampling.
Metadata
Related papers
Vibe Coding XR: Accelerating AI + XR Prototyping with XR Blocks and Gemini
Ruofei Du, Benjamin Hersh, David Li, Nels Numan, Xun Qian, Yanhe Chen, Zhongy... • 2026-03-25
Comparing Developer and LLM Biases in Code Evaluation
Aditya Mittal, Ryan Shar, Zichu Wu, Shyam Agarwal, Tongshuang Wu, Chris Donah... • 2026-03-25
The Stochastic Gap: A Markovian Framework for Pre-Deployment Reliability and Oversight-Cost Auditing in Agentic Artificial Intelligence
Biplab Pal, Santanu Bhattacharya • 2026-03-25
Retrieval Improvements Do Not Guarantee Better Answers: A Study of RAG for AI Policy QA
Saahil Mathur, Ryan David Rittner, Vedant Ajit Thakur, Daniel Stuart Schiff, ... • 2026-03-25
MARCH: Multi-Agent Reinforced Self-Check for LLM Hallucination
Zhuo Li, Yupeng Zhang, Pengyu Cheng, Jiajun Song, Mengyu Zhou, Hao Li, Shujie... • 2026-03-25
Raw Data (Debug)
{
"raw_xml": "<entry>\n <id>http://arxiv.org/abs/2603.19863v1</id>\n <title>MedQ-Engine: A Closed-Loop Data Engine for Evolving MLLMs in Medical Image Quality Assessment</title>\n <updated>2026-03-20T11:29:13Z</updated>\n <link href='https://arxiv.org/abs/2603.19863v1' rel='alternate' type='text/html'/>\n <link href='https://arxiv.org/pdf/2603.19863v1' rel='related' title='pdf' type='application/pdf'/>\n <summary>Medical image quality assessment (Med-IQA) is a prerequisite for clinical AI deployment, yet multimodal large language models (MLLMs) still fall substantially short of human experts, particularly when required to provide descriptive assessments with clinical reasoning beyond simple quality scores. However, improving them is hindered by the high cost of acquiring descriptive annotations and by the inability of one-time data collection to adapt to the model's evolving weaknesses. To address these challenges, we propose MedQ-Engine, a closed-loop data engine that iteratively evaluates the model to discover failure prototypes via data-driven clustering, explores a million-scale image pool using these prototypes as retrieval anchors with progressive human-in-the-loop annotation, and evolves through quality-assured fine-tuning, forming a self-improving cycle. Models are evaluated on complementary perception and description tasks. An entropy-guided routing mechanism triages annotations to minimize labeling cost. Experiments across five medical imaging modalities show that MedQ-Engine elevates an 8B-parameter model to surpass GPT-4o by over 13% and narrow the gap with human experts to only 4.34%, using only 10K annotations with more than 4x sample efficiency over random sampling.</summary>\n <category scheme='http://arxiv.org/schemas/atom' term='cs.CV'/>\n <published>2026-03-20T11:29:13Z</published>\n <arxiv:primary_category term='cs.CV'/>\n <author>\n <name>Jiyao Liu</name>\n </author>\n <author>\n <name>Junzhi Ning</name>\n </author>\n <author>\n <name>Wanying Qu</name>\n </author>\n <author>\n <name>Lihao Liu</name>\n </author>\n <author>\n <name>Chenglong Ma</name>\n </author>\n <author>\n <name>Junjun He</name>\n </author>\n <author>\n <name>Ningsheng Xu</name>\n </author>\n </entry>"
}