CMIG_paper_transformers_liver_MRI_release

Main changes:

• added training and hyperparameter tuning for sequence-specific and population encoder-only transformers using Pytorch and Optuna
• added files corresponding to preprocessed liver MRI dataset from OvGU
• added scripts related to statistical analysis of predictor performance
• improved overall quality

4th release

Added the following:

• time series prediction using kernel SVR
• DNI ablation experiment (disabling one of the terms in the gradient of ||f(A)||^2)
• improved visualization of signal resampling
• notebook with relative differences (performance) calculations to support the claims in the article

3rd release

Future frame prediction in 2D cine-MR images using the PCA respiratory motion model and online learning algorithms for RNNs, and time series forecasting using the latter.

Added scripts for visualization of multivariate time series forecasting with animated gifs

Second release

Future frame prediction in 2D cine-MR images using the PCA respiratory motion model and online learning algorithms for RNNs, and time series forecasting using the latter.

Added predicted video results to the repository, improved main readme, and added a second readme in the time series forecasting folder.

First release

v1.0.0

Added BSD-3-Clause license file