辅酶的主页

To code, to quote
and to note.

0%

Personal Introduction

selfie

English Version

My name is Jie (Tony) Feng, or Triple Coenzyme for network. I am a PhD student from Shanghai Jiao Tong University since 2020. Before that, I received my bachelor’s degree from Huazhong University of Science and Technology.

I had worked on cytopathological image super resolution and medical image semantic segmentation with deep learning. Now I focus on MRI research, including sequence coding, image reconstruction and radiomics analysis.

I have abundant experience of algorithm development, acceleration and deployment with python, matlab and C++. More project experience is shown below.

I am interested in Computer Vision and MRI reconstruction, especially combining physical model with deep learning to make CNN more constrained and explainable. I am always excited to contact with new technical improvement and hope to do some cool works in the future.

Education

2016.9 - 2020.6: Bachelor, Biomedical Engineering, Huazhong University of Science & Technology

2020.9 - present: PhD student, Biomedical Engineering, Shanghai Jiao Tong University

Skills

  • python
    • algorithm prototype development
    • algorithm acceleration with cython/numba
    • deep learning model training, deployment and quantization with pytorch/keras
    • GUI development with PyQt
  • matlab
    • algorithm prototype development
  • C++
    • UIH/Siemens MR sequence coding
    • multi-threaded high performance algorithm deployment with openmp
    • python wrapper development with pybind11
    • GUI development with Qt
  • docker
    • environment deployment
  • network and linux server management

Research Publications

  • Feng, J., Feng, R., … & Wei, H. (2025). Spatiotemporal implicit neural representation for unsupervised dynamic MRI reconstruction. IEEE transactions on medical imaging, Early Access.

  • Feng, J., Chen, J., … & Wei, H. (2025). Calibration-free DCE-MRI with sub-second temporal resolution using interpretable implicit neural representation. ISMRM Abstract(Oral), 4345.

  • Zhang, M., Feng, R., Li, Z., Feng, J., … & Wei, H. (2024). A subject-specific unsupervised deep learning method for quantitative susceptibility mapping using implicit neural representation. Medical Image Analysis, 95, 103173.

  • Feng, R., Wu, Q., Feng, J., … & Wei, H. (2023). IMJENSE: scan-specific implicit representation for joint coil sensitivity and image estimation in parallel MRI. IEEE transactions on medical imaging, 43(4), 1539-1553.

  • Chen, M., Wang, Y., Shi, Y., Feng, J., … & Wei, H. (2023). Brain Age Prediction Based on Quantitative Susceptibility Mapping Using the Segmentation Transformer. IEEE Journal of Biomedical and Health Informatics, 28(2), 1012-1021.

  • Li, Z., Feng, R., Liu, Q., Feng, J., … & Wei, H. (2023). APART-QSM: An improved sub-voxel quantitative susceptibility mapping for susceptibility source separation using an iterative data fitting method. Neuroimage, 274, 120148.

  • Feng, J., Zhang, M., Feng, R. & Wei, H. (2023). SSMo-QSM: A self-supervised learning method for model-based quantitative susceptibility mapping reconstruction. ISMRM Abstract, 3099.

  • Wu, Y., Zhang, C., Li, Y., Feng, J., … & Sun, B. (2022). Imaging insights of isolated idiopathic Dystonia: Voxel-based morphometry and activation likelihood estimation studies. Frontiers in Neurology, 13, 823882.

  • Feng, R., Zhao, J., Wang, H., Yang, B., Feng, J., … & Wei, H. (2021). MoDL-QSM: Model-based deep learning for quantitative susceptibility mapping. NeuroImage, 240, 118376.

  • Ma, J., Yu, J., Liu, S., Chen, L., Li, X., Feng, J., … & Cheng, S. (2020). PathSRGAN: Multi-supervised super-resolution for cytopathological images using generative adversarial network. IEEE transactions on medical imaging, 39(9), 2920-2930.

Personal Project Experience

This part only contains the projects leaded or finished individually by Jie Feng.

  • Vertebra Segmentation in MRI with ResNet-Unet (Github link)
    • Winner of 2019 National Undergraduate Biomedical Engineering Innovation Design Competition.

  • MRI Motion Correction with Patch-GAN (Github link)

  • RealBrain Software (https://qmri.sjtu.edu.cn/resources), a PyQt-based DBS (Deep Brain Stimulation) Surgery Planning System, including:

    • mature GUI
    • multi-modality medical image IO / registration
    • CT frame extraction
    • QSM (Quantitative Susceptibility Mapping) automatic reconstruction
    • QSM-based DGN (Deep Brain Nuclei) accurate segmentation
    • DTI reconstruction and fiber tracking
    • surgery target localization and path planning
  • Real-time MR Imaging Pipeline during DBS surgery, including:
    • a customized golden-angle radial sequence in siemens scanner
    • an external GPU-based real-time reconstruction with CNN (Github link)
  • Standalone Python Wrapper for Brain Extraction Tool (bet2) (Github link)

Language

  • Chinese
  • English
  • a little bit Japanese

中文版本

我叫冯颉，网名辅酶的辅酶的辅酶，目前是上海交通大学的一名博士生，本科毕业于华中科技大学生医学工程系。

我曾做过基于深度学习的病理图像超分辨以及医学图像分割相关的工作，目前则主要从事磁共振研究的相关工作，包括磁共振序列编写、图像重建以及影像组学分析等。

我能比较熟练地使用python、matlab复现、开发、优化各类算法，并能够使用C++进行部署。具体可见下面的个人项目经历部分。

我目前的研究兴趣主要在于计算机视觉以及磁共振图像重建，尤其是通过物理模型更好约束神经网络、提升其可解释性方面。我对任何的技术新进展都充满好奇并乐于尝试，更希望运用它们在未来做出一些有意思的工作。

教育背景

2016.9 - 2020.6: 华中科技大学，生物医学工程，学士

2020.9 - 目前: 上海交通大学， 生物医学工程，博士生

技能

  • python
    • 算法开发
    • 基于cython/numba的算法加速
    • 基于pytorch/keras的深度学习模型训练、部署及量化
    • 基于PyQt的图形界面开发
  • matlab
    • 算法开发
  • C++
    • 联影、西门子磁共振序列编写
    • 基于openmp的多线程算法加速
    • 基于pybind11的python接口编写
    • 基于Qt的图形界面开发
  • docker
    • 环境部署
  • 网络及linux服务器的管理与维护

发表论文

  • Feng, J., Feng, R., … & Wei, H. (2025). Spatiotemporal implicit neural representation for unsupervised dynamic MRI reconstruction. IEEE transactions on medical imaging, Early Access.

  • Feng, J., Chen, J., … & Wei, H. (2025). Calibration-free DCE-MRI with sub-second temporal resolution using interpretable implicit neural representation. ISMRM Abstract(Oral), 4345.

  • Zhang, M., Feng, R., Li, Z., Feng, J., … & Wei, H. (2024). A subject-specific unsupervised deep learning method for quantitative susceptibility mapping using implicit neural representation. Medical Image Analysis, 95, 103173.

  • Feng, R., Wu, Q., Feng, J., … & Wei, H. (2023). IMJENSE: scan-specific implicit representation for joint coil sensitivity and image estimation in parallel MRI. IEEE transactions on medical imaging, 43(4), 1539-1553.

  • Chen, M., Wang, Y., Shi, Y., Feng, J., … & Wei, H. (2023). Brain Age Prediction Based on Quantitative Susceptibility Mapping Using the Segmentation Transformer. IEEE Journal of Biomedical and Health Informatics, 28(2), 1012-1021.

  • Li, Z., Feng, R., Liu, Q., Feng, J., … & Wei, H. (2023). APART-QSM: An improved sub-voxel quantitative susceptibility mapping for susceptibility source separation using an iterative data fitting method. Neuroimage, 274, 120148.

  • Feng, J., Zhang, M., Feng, R. & Wei, H. (2023). SSMo-QSM: A self-supervised learning method for model-based quantitative susceptibility mapping reconstruction. ISMRM Abstract, 3099.

  • Wu, Y., Zhang, C., Li, Y., Feng, J., … & Sun, B. (2022). Imaging insights of isolated idiopathic Dystonia: Voxel-based morphometry and activation likelihood estimation studies. Frontiers in Neurology, 13, 823882.

  • Feng, R., Zhao, J., Wang, H., Yang, B., Feng, J., … & Wei, H. (2021). MoDL-QSM: Model-based deep learning for quantitative susceptibility mapping. NeuroImage, 240, 118376.

  • Ma, J., Yu, J., Liu, S., Chen, L., Li, X., Feng, J., … & Cheng, S. (2020). PathSRGAN: Multi-supervised super-resolution for cytopathological images using generative adversarial network. IEEE transactions on medical imaging, 39(9), 2920-2930.

个人项目经历

此部分仅包含本人作为负责人或独立完成的项目。

  • 基于ResNet-Unet的磁共振图像脊柱分割 (Github 链接)
    • 本项目获得第五届全国大学生生物医学工程创新设计竞赛命题组一等奖

  • 基于Patch-GAN的磁共振运动矫正 (Github 链接)

  • RealBrain软件 (https://qmri.sjtu.edu.cn/resources), 是一个基于PyQt开发的可视化脑深部刺激（DBS, Deep Brain Stimulation）手术规划系统，其具有以下功能:

    • 成熟的图形界面
    • 多模态医学图像的导入、配准
    • 基于CT图像的自动头架提取
    • 定量磁化率成像（QSM, Quantitative Susceptibility Mapping）全自动重建
    • 基于定量磁化率成像的脑深部核团精准分割
    • DTI图像重建及纤维追踪可视化
    • 可视化手术靶点定位及手术路径规划
  • 脑深部刺激手术过程中的实时成像流程, 包括以下工作:
    • 西门子磁共振扫描仪中可用的自定义黄金角径向实时成像序列
    • 基于GPU的实时重建流水线，可加入神经网络(Github 链接)

语言

  • 中文
  • 英文
  • 一点点日文