首页» 教师队伍» 导师介绍

导师介绍

万沛星 研究员


万沛星  研究员 助理教授

研究方向:医学大模型研发与临床应用

peixing@bjmu.edu.cn


教育经历:

20109-20157  武汉大学临床医学专业   获医学学士

20159-20207  中山大学临床医学专业   获医学博士

201810-20211 密歇根大学生物医学专业  联合培养


工作经历:

20211-20252 美国国立卫生研究院National Institutes of Health  博士后

20253-至今 北京大学基础医学院医学生物信息学系  Tenure-track助理教授


个人简介:

国家级高层次青年人才计划入选者。专注医疗人工智能交叉领域,具备扎实的临床背景与领先的科研能力,在医疗对话大模型的多项关键技术和临床验证方面实现原创性突破。在武汉大学和中山大学接受系统临床医学培养,并在美国密歇根大学与国立卫生研究院(NIH)完成高水平科研训练,形成了融合医学、人工智能与临床转化的多学科研究优势。持续产出具有国际影响力的成果,以通讯/第一作者在Nature MedicineMedCell Reports Medicinenpj Digital MedicineCell Genomics (共一)发表原创性论文;在LancetLancet Planetary Health Clinical and Translational Medicine等期刊发表观点述评;Nature Medicine研究成果入选同期专题报道;三篇高影响力论文引用数超过300次。


研究方向:

医疗对话大模型在信息输入和决策透明度上仍面临诸多挑战。一方面,当前模型主要依赖文本数据,难以完整表征临床决策所需的多维度信息,如语气、表情、肢体语言等非语言信号。另一方面,现有模型的推理过程缺乏透明度,难以提供清晰、可解释的决策依据,既增加了临床风险管控难度,又削弱了医患双方对大模型诊疗建议的信任基础。针对上述挑战,设以下三个研究方向:

方向一:多模态医疗大模型构建

方向二:提升医疗大模型决策可解释性

方向三:医疗大模型落地模式临床试验


代表性论文

1. Outpatient reception via collaboration between nurses and a large language model: a randomized controlled trial.

Wan P*, Huang Z*, Tang W, Nie Y, Pei D, Deng S, Chen J, Zhou Y, Duan H, Chen Q#, Long E#.

Nature Medicine. 2024 (Highlighted by Nature Medicine, ‘Improving primary healthcare with generative AI’)

2. A bimodal large language model reduces misalignment in patient education: A double-blinded randomized trial.

Wan P*#, Huang Z*, Huang H*, Liang G*, …, Cao M#, Long E#.

Med. 2026

3. From raw audio to structure: an agent-based pipeline that boosts medical LLM performance.

Qin H*, Tang W*, Huang Z*, Liang G*, …, Long E#, Wan P#

npj Digital Medicine. 2026

4. Model confrontation and collaboration: A debate intelligence framework for enhancing medical reasoning in large language models.

Sun X*, Hong Q*, …, Wan P#, Long E#

Cell Reports Medicine. 2026

5. From function to translation: Decoding genetic susceptibility to human diseases via artificial intelligence.

Long E#*, Wan P*, Chen Q, Lu Z, Choi J#.

Cell Genomics. 2023

6. Soluble tissue factor generated by necroptosis-triggered shedding is responsible for thrombosis.

Wan P*, Choksi S*, Park Y*, Chen X*, Yan J, Foroutannejad S, Liu Z, Chen J, Lake R, Liu C, Liu Z#.

Cell Research. 2025

7. LncRNA H19 initiates microglial pyroptosis and neuronal death in retinal ischemia/reperfusion injury.

Wan P*, Su W*, Zhang Y, Li Z, Deng C, Li J, Jiang N, Huang S, Long E, Zhuo Y#.

Cell Death & Differentiation. 2020

8. TET-dependent GDF7 hypomethylation impairs aqueous humor outflow and serves as a potential therapeutic target in glaucoma.

Wan P*, Long E*, Li Z, Zhu Y, Su W, Zhuo Y#.

Molecular Therapy. 2021

9. From General to Specific: Tailoring Large Language Models for Real-World Medical Communications.

Sun X*, Tang W*, …, Long E#, Wan P#.

Clinical and Translational Medicine. 2025

10. Expectations of medical students in China.

Wan P, Long E#.

The Lancet. 2016

11. Use of a Neural Net to Model the Impact of Optical Coherence Tomography Abnormalities on Vision in Age-related Macular Degeneration.

Wan P, Long E#.

American Journal of Ophthalmology. 2018

12. Predicting Real-World Future of Glaucoma Patients? Cautions Are Required for Machine Learning.

Long E, Wan P#, Zhuo Y.

Translational Vision Science and Technology. 2017



名片.png