Dr. Sifan Song is an Assistant Professor at Xi’an Jiaotong-Liverpool University. Prior to joining XJTLU, he served as a Research Fellow at Harvard Medical School and Massachusetts General Hospital (2024–2026), working with Dr. Quanzheng Li, Dr. Xiang Li, and Dr. Dufan Wu. His research focuses on representation learning, parameter-efficient fine-tuning, and explainable frameworks for large-scale natural and medical images, with an emphasis on 2D/3D segmentation and generation across CT, MRI, and PET, as well as efficient multimodal learning and the integration of large language models and multi-agent systems for real-world applications.

He received a PhD in Computer Science from the University of Liverpool. He also holds an MSc in Artificial Intelligence from the University of Southampton and an MSc in Bioinformatics from the University of Edinburgh. During his doctoral training, he was a Visiting Doctoral Student with the MIRACLE team at the University of Science and Technology of China (2022–2024), working with Prof. Shaohua (Kevin) Zhou, and a Visiting Doctoral Student at The Chinese University of Hong Kong (Shenzhen) (2019–2020), working with Dr. Jiaming Zhang.

He serves as a reviewer for IEEE Transactions on Medical Imaging, IEEE Transactions on Artificial Intelligence, and Medical Physics, and regularly reviews for conferences such as AAAI, CVPR, ICLR, NeurIPS, MICCAI, ECCV, WACV, and ISBI. Dr. Song has also collaborated with GE HealthCare on projects and clinical pipelines, and has worked with medical AI companies. His current work aims to bridge advanced AI methodologies with real-world problems and interdisciplinary education.

Assistant Professor, Xi'an Jiaotong-Liverpool University, 2026.02 ~ present

Resarch Fellow, Harvard Medical School and Massachusetts General Hospital, 2024.02  ~ 2026.02

My office is at D-5077, and you are welcome to visit me during my office hours (Monday 15:00–17:00 (XEC D-5077), Thursday 13:00–15:00 (SIP CB144)), or send me an email to make an appointment.

AY25/26-S2: Co-teacher of DTS002, DTS004

• Best Paper Award Finalist (ISBI 2022): Bilateral-vit for robust fovea localization.
• 2nd Place in MICCAI 2021 Contest, GAMMA Challenge (2nd in sub-task 1, glaucoma grading based on 3D multi-modality images; 1st in sub-task 3, optic disc and cup segmentation)