We are a research laboratory led by Professor Feng Gao at the Sixth Affiliated Hospital of Sun Yat-sen University in Guangzhou, China. We bring together clinical expertise and computational research to develop explainable, multimodal artificial intelligence for cancer research and clinical care, with colorectal cancer as a principal focus.
Our work spans clinical reasoning and evidence-based AI agents, multimodal learning across imaging and molecular data, computational methods for surgical care, and interpretable models of disease mechanisms.
Lab website · Research · Publications · People
This organization hosts the lab's new research code and software releases and provides an index of projects presented on the lab website. Earlier projects remain available in their original repositories, preserving their publication links and project-specific licenses. Code links below lead to verified public repositories; availability of code does not imply that every experiment, dataset, or model weight is included.
| Project | Research focus | Resources |
|---|---|---|
| Brim | Interpretable fusion of histology and molecular data for pan-cancer prognosis. | Code · Paper · Project |
| CRCFound | A self-supervised CT foundation model for colorectal cancer diagnosis, staging, molecular prediction, and prognosis. | Code · Paper · Project |
| PIANOS | A normalization-free, single-sample classifier for colorectal cancer risk stratification across platforms. | Code · Paper · Project |
| SegMamba-V2 | Long-range volumetric modeling for 3D medical image segmentation across organs and modalities. | Code · Paper · Project |
| TMO-Net | Explainable multi-omics pretraining for transferable prediction and pathway-level interpretation in oncology. | Code · Paper · Project |
| SOUSA | Weakly and semi-supervised medical image segmentation using sparse annotations and unlabeled images. | Paper · Project |
| 3D RP-Net | Longitudinal multi-task MRI modeling for treatment response prediction in rectal cancer. | Paper · Project |
| DeepCC | Pathway-informed deep learning for cancer molecular subtype classification. | Code · Paper · Project |
For each software project, consult its README for installation, data access, reproducibility scope, and citation instructions. Where no code link is listed, consult the project page or publication for availability information.
- ICGC-ARGO CRC: a colorectal cancer clinical genomics program supporting translational research and biomarker development. Consult the project page for program and access information.
- One Person Lab App: a local-first workbench for expert knowledge work and research agent workflows. Source and releases.
- OPL Cloud: a cloud platform for coordinating projects, resources, and research agent services. Source · Current status.
The lab research portfolio provides further context and related publications.
We welcome research collaborations, student inquiries, and contributions to our software. For questions about a particular tool or a reproducibility issue, please use that repository's Issues and consult our contribution guidance.
For research collaborations, prospective student inquiries, or lab visits, contact Feng Gao or visit the lab's contact page. Current team information is available on the People page.
Start with the research repository template and the Research Software Publishing Guide. Members can create repositories and collaborate directly; templates and review practices can be adapted to each project.
New lab open-source code and packages use Apache License 2.0 by default, with documentation in professional academic English. Each repository identifies its authors, maintainers, license, and publication-specific versions. Data and model licenses are stated separately. Patient-level data must not be uploaded to GitHub, including private repositories.
