Computational Pathology
计算病理与全切片图像分析
Weakly supervised and interpretable learning for whole-slide pathological images under real-world small-sample settings.
面向真实世界小样本条件下的全切片病理图像弱监督学习、患者参照建模与可解释性分析。
李氏医学人工智能实验室 · Biomedical AI Lab
Li-lab
Medical AI for Real-world Biomedical Discovery
Li-lab develops interpretable, data-efficient, and clinically grounded AI methods for computational pathology, multimodal biomedical data integration, and translational medical AI.
Li-lab 致力于发展可解释、数据高效且面向真实临床场景的医学人工智能方法，重点关注计算病理、多源生物医学数据整合与转化医学人工智能。
Computational pathology × Real-world dataFrom WSI patches to clinically grounded, interpretable AI models.
Research
Research areas
We focus on methods that are robust under limited labels, domain shift, and real clinical heterogeneity.
Biomedical Data Integration
多源生物医学数据整合
Integrating pathology images, molecular profiles, clinical variables, and public biomedical resources for robust translational modeling.
整合病理图像、分子特征、临床变量与公共生物医学资源，构建稳健的转化医学建模框架。
Interpretable and Data-efficient AI
可解释与数据高效人工智能
Developing prototype-guided, domain-adaptive, and uncertainty-aware AI models for limited-label biomedical applications.
发展原型引导、领域适配和不确定性建模方法，服务于有限标注条件下的生物医学智能分析。
Projects
Featured projects
Representative projects are organized as reusable cards, so future students only need to edit data files.
PRISM-MIL
Patient-referenced interpretable stabilized MIL for WSI analysis
A weakly supervised framework that uses paired normal mucosa as patient-specific reference for real-world small-sample colorectal whole-slide image analysis.
M4DR
Multimodal modeling for drug response prediction
A multimodal framework integrating drug features, cellular profiles, pathway signals, and cross-domain transfer learning for precision oncology modeling.
Perinatal Mental Health AI
Evidence synthesis and psychosocial risk modeling
Quantitative evidence synthesis and statistical modeling for perinatal mental distress, parent-infant bonding, and psychosocial protective factors.
People
Team members
Li-lab welcomes students from medicine, computer science, statistics, bioinformatics, and biomedical engineering.
Prof. Ming Li · 李明 教授
Principal Investigator · 课题组负责人 / 教授 / 博士生导师
Computational pathology, weakly supervised learning, interpretable medical AI, and real-world clinical data science.
计算病理、弱监督学习、可解释医学人工智能与真实世界临床数据科学。
Prof. Ming Li
Principal Investigator
Taylor Wang
MSc Student
Yue Chen
MSc Student
Hao Lin
Undergraduate Researcher
Publications
Selected publications
Replace these placeholder papers with your real manuscripts, journal articles, preprints, and conference papers.
2026
Patient-referenced weakly supervised adaptation for real-world small-sample gastrointestinal WSI analysis
2026
Multimodal Four-Tower Drug Response Modeling with Cross-domain Representation Alignment
2025
Network-based psychosocial profiling of stigma and fertility-related stress in reproductive health research
News
Latest news
News makes the lab look active. Keep it short, dated, and easy to update.
2026-06-22
Li-lab website project started
2026-06-15
New computational pathology project launched
2026-06-01
New real-world WSI cohort organized
2026-05-20
Weekly journal club resumed
Join Us
Work with Li-lab
We are looking for motivated students interested in medical AI, computational pathology, weakly supervised learning, and biomedical data science.
欢迎对医学人工智能、计算病理、弱监督学习和真实世界医学数据建模感兴趣的同学加入。
Email usContact
Email: li-lab@example.edu.cn
Address: School of Biomedical Engineering, Example Medical University
示例医科大学 生物医学工程学院
City: Haikou, China