Giters

sravan953 / delors

Application of Deep Learning for Rapid Reconstruciton of Spiral K-Space data for Magnetic Resonance Imaging

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

DeLorRS

Application of Deep Learning Neural Networks for Rapid Reconstruction of Spiral K-Space Data for Magnetic Resonance Imaging

Project submission for Intro to ML (EL-9123) coursework.

Setup

  1. Clone repo
  2. Download and generate data (read below)
  3. Run network

Downloading data

  1. Clone repo
  2. Download grayscale images from ImageNet: python3 get_data_imagenet.py in your terminal
  3. Run IterAddPhase.m in Matlab on all the downloaded folders
  4. Download GPI v1-beta, and install
  5. Generate k-space data using samples from step 3 and complex_spiral_traj280_58.mat in the gen_kspace.net
  6. Place all the k-space data (in their belonging folders) in imagenet_64_kspace and the original images (in their belonging folders) in imagenet_64_orig

Contributing

Fork and PR!

About

Application of Deep Learning for Rapid Reconstruciton of Spiral K-Space data for Magnetic Resonance Imaging

License:MIT License

Languages

Language:Jupyter Notebook 82.4%Language:Python 10.1%Language:MATLAB 7.6%