Reliable CT reconstruction
Low-dose and sparse-view reconstruction, super-resolution, and artifact correction with physical and data-driven priors.
Low-dose CTMARUncertainty
Chief Scientist / JancsiTech / Shenzhen, China
Reliable AI
for medical imaging.
I develop methods for CT reconstruction, artifact reduction, and medical image synthesis, and study their use in imaging-based cancer assessment.
MEDICAL AI / 01
CTRECONSTRUCTION
From sparse-view acquisition
to robust clinical inference.
to robust clinical inference.
2,567Google Scholar citations
22h-index
30i10-index
85+published papers
Google Scholar snapshot: October 2026
Research program
Making medical imaging AI more reliable, useful, and reusable.
My work follows the imaging pipeline from imperfect acquisition to clinically meaningful prediction.
Medical image synthesis
Diffusion and multimodal models for CBCT-to-CT translation, virtual imaging, and controllable medical image generation.
DiffusionCBCT-to-CTMRI
Imaging-based oncology
Imaging and pathomics models for recurrence, immune context, treatment response, and disease assessment.
PrognosisImmunotherapyPathomics
Open-source work
Methods that others can read, run, and build on.
Research code and paper-linked implementations, organized as one reproducible entry point for collaborators and readers.
01IEEE TMI / 2018
DD-Net
A sparse-view CT reconstruction method based on combination of DenseNet and deconvolution.
DenseNet and deconvolution for sparse-view reconstruction. ESI Highly Cited Paper.
02MICCAI / 2021
TransCT
Dual-path Transformer for low-dose computed tomography.
A transformer-based reconstruction path for low-dose CT with self-supervised learning.
03IEEE TMI / 2020
Deep sinogram completion
Image-prior guided metal artifact reduction in CT images.
Sinogram completion paired with image priors for robust metal artifact reduction.
Paper ↗206 citations
04MIA / 2025
TPDM-CBCT2CT
Texture-preserving diffusion model for CBCT-to-CT synthesis.
Diffusion-based image translation designed to preserve texture for downstream imaging tasks.
Additional repositories
More public code connected to imaging, segmentation, and oncology studies.
01 / diffusion MRIDoDTI
Reliable deep diffusion tensor estimation with data-driven optimization.
GitHub ↗ 02 / endoscopyEC-DLDeep learning for endocytoscopic classification of colorectal lesions.
GitHub ↗ 03 / segmentationFedU-DAFederated multi-organ segmentation with dynamic attention.
GitHub ↗ 04 / oncologyClassGastricBiology-guided prognosis and immunotherapy response prediction.
GitHub ↗ 05 / gastric cancerPR_survival_PredictionCT-based prediction of peritoneal recurrence and disease-free survival.
GitHub ↗Clinical applications
From better images to better questions.
Selected publications
Recent work and high-impact contributions.
CBCT→CTdiffusion translation
Texture-preserving diffusion model for CBCT-to-CT synthesisPaper ↗
endocytoscopic classification
Deep learning based endocytoscopic classification for colorectal lesionsPaper ↗
projection→DenseNetsparse-view reconstruction
A sparse-view CT reconstruction method based on DenseNet and deconvolutionPaper ↗
image priorsinogram completion
Deep sinogram completion with image prior for metal artifact reductionPaper ↗
CT+survivalmultitask learning
Predicting peritoneal recurrence and disease-free survival from CT imagesPaper ↗
image→immune contextbiology-guided learning
Biology-guided deep learning for prognosis and immunotherapy responsePaper ↗
About
Research leadership across academia and industry.
I am Chief Scientist at JancsiTech, where I lead medical imaging AI research and development. I previously worked as a Research Leader in ByteDance AI Healthcare and as a postdoctoral research fellow at Stanford University with Prof. Lei Xing.
My Ph.D. was completed at the Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, and the University of Chinese Academy of Sciences.
2022 - presentJancsiTechChief Scientist
2020 - 2022ByteDance AI HealthcareResearch Leader
2019 - 2022Stanford UniversityPostdoctoral Research Fellow
2017 - 2018Virginia TechVisiting Scholar
Top 2% Scientists 2025-2026 ↗