Clinically Reliable Medical Image Analysis

I am a PhD student in Biomedical Engineering at Shanghai Jiao Tong University, jointly supervised by Prof. Jianqi Sun and Associate Professor Xia Li. I began the direct-entry doctoral programme in 2023 after completing my undergraduate degree in biomedical engineering at the same school. My work sits at the intersection of medical-image computing and clinically grounded quantitative analysis.

My main research concerns cervical spine imaging. I study pathology-guided segmentation and measurement in cervical MRI, uncertainty-aware detection of spinal-cord T2 hyperintensity, and reconstruction of isotropic three-dimensional cervical MRI from anisotropic routine scans. In CT, I have worked on pseudo-healthy vertebral synthesis and relative height-loss measurement for interpretable assessment of compression fractures. Across these projects, anatomical structure is not merely an input: it also provides a way to inspect model outputs and relate them to quantitative imaging indicators.

I also develop unsupervised anomaly-detection methods for settings in which lesion annotations are scarce or incomplete. This research has progressed from cross-sequence translation to a unified framework designed for varied brain diseases, MRI sequences, and missing-input settings. The methods combine content–style disentanglement, image translation, reconstruction evidence, feature-level comparison, and uncertainty estimation. I am particularly interested in how these components behave under changes in scanners, acquisition protocols, and clinical populations.

Clinical collaboration shapes the questions I choose and the way I evaluate a method. I currently work with teams in spine surgery, orthopaedics, and radiology, including Renji Hospital, Shanghai Sixth People’s Hospital, and Ruijin Hospital. My aim is to develop methods that are evaluated beyond a single development dataset and whose results can be reviewed through anatomy, uncertainty, and measurable imaging evidence.