Junfeng Jiang

Biography

Junfeng Jiang is an Associate Professor at the College of Artificial Intelligence and Automation, Hohai University. He serves as the Deputy Director of the Institute of Human–Environment Intelligent Interaction and Systems.

His research focuses on AI-driven medical treatment planning and execution, orthopedic surgical robotics, medical image analysis, and 3D computer graphics. His work has been published in leading venues including Medical Physics, ICCV, MICCAI, EuroGraphics, Computer-Aided Design, and Computers & Industry. In 2026, his group published a semantic edge-guided single-view vertebral registration method in Medical Physics and had FlowReg published at MICCAI 2026; an extended radiative Gaussian registration framework, RadGS-Reg+, was published online in Medical Image Analysis.

Research Areas

• AI Medical Treatment Planning and Execution
• Orthopedic Surgical Robotics
• Medical Image Analysis
• 3D Computer Graphics
• Fracture Reconstruction and Reduction

Selected Publications

Ao Shen, Junfeng Jiang*, Xueming Fu, Qiang Zeng, Ye Tang, Zhengming Chen, Luming Nong, Feng Wang, Shaohua Kevin Zhou*
FlowReg: Reconstruction-Guided Flow Matching for Universal Spinal CT/X-Ray Registration

MICCAI 2026 · Published online 2026-09-25 · LNCS 16889, pp. 77–87

DOI Official paper Repository

FlowReg combines biplanar X-ray reconstruction with flow-based pose estimation for both single- and multi-vertebra CT/X-ray registration. Reconstructed volumes provide spatial guidance, while SE(3)-aware supervision and difference-aware feature fusion support pose learning and cross-modal alignment.

* Joint corresponding authors. Code forthcoming.

RadGS-Reg+: Teacher-Student Radiative Gaussian Reconstruction and 3D/3D Registration for Lumbar CT/X-ray Registration

Junfeng Jiang†*, Ao Shen†, Zhan Wang, Xueming Fu, Zhengming Chen, Luming Nong, Feng Wang, Shaohua Kevin Zhou*

Medical Image Analysis · Online 2026 · Volume 115 (2027), Article 104319

DOI

RadGS-Reg+ reconstructs a registration-oriented vertebral representation from biplanar X-rays and aligns it with preoperative CT in 3D. Building on RadGS-Reg, it adds CT-to-X-ray teacher–student distillation, adaptive biplanar feature fusion, and discrepancy-aware pose regression.

† Joint first authors. * Joint corresponding authors.

Shen A, Jiang J*, Tang Y, Chen Z, Nong L, Chen Q, Wang F, Wuse X
Semantic Edge-Guided Single-View 2D/3D Registration for Vertebrae in X-rays

Medical Physics · 2026

Abstract

A single-view 2D/3D vertebra registration method guided by semantic edge information, designed to improve robustness and geometric consistency in X-ray based spinal alignment tasks.

Ao Shen, Xueming Fu, Junfeng Jiang*, Qiang Zeng, Ye Tang, Zhengming Chen, Luming Nong, Feng Wang, S. Kevin Zhou*
RadGS-Reg: Registering Spine CT with Biplanar X-Rays via Joint 3D Radiative Gaussians Reconstruction and 3D/3D Registration

MICCAI 2025 · LNCS 15968, pp. 456–466

Best Paper Nomination Young Scientist Award Nomination

Paper DOI Official paper Project Code

Jointly reconstructs a 3D radiative Gaussian representation from biplanar X-rays and registers it with preoperative CT. Counterfactual attention focuses learning on vertebral regions, while patient-specific pretraining supports adaptation from simulated to intraoperative images.

* Joint corresponding authors. Recognition: Best Paper Nomination; Young Scientist Award Nomination.

Abstract

A framework for spine CT and biplanar X-ray registration that combines 3D radiative Gaussian reconstruction with 3D/3D registration to improve alignment accuracy in image-guided spinal intervention scenarios.

Lu J, Liang Y, Han H, Jiang J*, et al.
A Survey on Computational Solutions for Reconstructing Complete Objects by Reassembling Their Fractured Parts

Computer Graphics Forum · 2025

Paper

Abstract

A survey of computational methods for reconstructing complete objects from fractured parts, covering geometric matching, learning-based assembly, optimization strategies, datasets, and evaluation protocols.

Wang Z, Jiang J*, et al.
Spinal Magnetic Resonance Imaging Super-Resolution Based on Improved Latent Diffusion Models

2025 2nd International Conference on Image, Signal Processing and Communication Technology (ISPCT), IEEE, 2025: 206–212

Abstract

A spinal MRI super-resolution method based on improved latent diffusion models, designed to enhance image quality and anatomical detail for downstream analysis and clinical applications.

Jiang J*, Shuai L, Zeng Q
ABLSpineLevelCheck: Localization of Vertebral Levels on Fluoroscopy via Semi-supervised Abductive Learning

IEEE BIBM 2024

Abstract

A semi-supervised abductive learning framework for localizing vertebral levels on fluoroscopic images, targeting more reliable intraoperative level identification in spinal procedures.

Deng Z, Jiang J*, Chen Z, et al.
TAssembly: Data-driven Fractured Object Assembly

Computers & Graphics · 2023

Abstract

A data-driven framework for fractured object assembly using a linear template model to guide fragment alignment and reconstruction.

Deng Z, Jiang J*, Huang R, et al.
Synergistically Segmenting and Reducing Fracture Bones via Whole-to-Whole Deep Dense Matching

Computers & Graphics · 2023

Abstract

A whole-to-whole deep dense matching approach that jointly performs fracture bone segmentation and reduction to support intelligent orthopedic reconstruction.

Wang F, Jiang J*, Deng Z, et al.
A Deep-Learning Approach for Locating the Intramedullary Nail's Holes Based on 2D Calibrated Fluoroscopic Images

Journal of Mechanics in Medicine and Biology · 2023

Abstract

A deep-learning method for localizing intramedullary nail holes from calibrated fluoroscopic images, aiming to improve efficiency and accuracy in orthopedic fixation procedures.

Huang Q, Huang X, Sun B, Jiang J, et al.
ARAPReg: An As-Rigid-As-Possible Regularization Loss for Learning Deformable Shape Generators

ICCV 2021

Paper Code

Abstract

An as-rigid-as-possible regularization loss for learning deformable shape generators, encouraging plausible deformation behavior while preserving local geometric structure.

* Corresponding author

Research Projects & Grants

Ongoing Projects

Completed Projects

Software

Students & Alumni

• PhD: Ziyue Deng — Huawei 2012 Laboratory
• PhD: Ao Shen — Postdoctoral Researcher at Shanghai Jiao Tong University
• Master graduates employed at NIO, ByteDance, Meituan, banks, and state-owned enterprises

Teaching

• Machine Learning — Fall Semester (Annual)
• AI Medical Image Analysis and Applications — Spring Semester (Annual)

Awards & Honors

• MICCAI 2025 Best Paper Nomination
• MICCAI 2025 Young Scientist Award Nomination
• Jiangsu Medical Science Award, 2023
• Students advised by the group have received Outstanding Master’s Thesis honors in Jiangsu Province