Zheyu Wen is a PhD student in University of Texas at Austin.
His interests include Computational medical image, Spatial temporal dynamic modeling, High performance computing and Machine Learning.
I graduated from University of Electrical and Scientific Technology of China with BS in Communications Engineering in 2019.
Then, I completed the my M.S. degree in Electrical and Computer Engineering in 2021 in University of Michigan, Ann Arbor.
Research experience
Computational modeling, inversion for Alzheimer's disease
Jun. 2022 -- Now
Proposed new model describing pathological tau spreading.
Publication on MICCAI workshop MTSAIL+LEAF 2023.
On going work:
Solve ODE and sparsity constrained optimization problem with Newton method and Gradient informed sparsity.
Optimize a high dimensional variable using C++ CUDA and Parallel MCMC.
Construct brain connectivity matrix by CUPY and Joblib, and publish as a open software.
Representation learning of resting state fMRI with VAE
May 2020 -- Sep. 2021
Compressed high-dimensional fMRI image into a low dimension vector.
Publication on NeuroImage, 2021
Implement VAE by Pytorch focusing on 3D fMRI image.
Design a spatio-temporal dynamic model based on Neural ODE
Unrolling-net MRI reconstruction
Jun. 2021 -- Nov. 2021
Inspired from iterative inversion algorithm.
Unrolling-net iteratively call network for denoising, and achieves SOTA.