This repository contains the source code of the paper Rician noise estimation for 3D Magnetic Resonance Images based on Benford's Law (URL doi paper https://doi.org/10.1007/978-3-030-87231-1_33.
This project consist of 3 directories:
- input: it contains 2 3D-MRI brain images form the NKI-RS repository.
- output: the folder where the results will have save it.
- src: the code consist in 2 Jupyter files:
- 01_loadMRI_boxplot.ipynb
- 02_randomData_method.ipynb
Easily you can download the code and ejecute each Jupyter to visualize:
Files obtained from the '01_loadMRI_boxplot.ipynb' file:
- 1 csv file:
- 10 equal spaced noise added to each MRI, Bhattacharyya Coefficient and Kullback.Leibler divergence values for each noise added.
- 2 boxplots, to represent the values of each colum of the csv above:
- Bhattacharyya Coefficient vs noise.
- Kullback.Leibler divergence vs noise.
Files obtained from the '02_randomData_method.ipynb' file:
- 1 csv file:
- 20 random noise added to each MRI, Bhattacharyya Coefficient and Kullback.Leibler divergence values for each noise added.
- 2 tables with MSE value for each regressor:
- Bhattacharyya Coefficient.
- Kullback.Leibler divergence.
- 2 tables with R2 value for each regressor:
- Bhattacharyya Coefficient.
- Kullback.Leibler divergence.
- 2 scatterplot with the regressors:
- Bhattacharyya Coefficient.
- Kullback.Leibler divergence.
The content you should be able to see is in the 'output_should_be_appear' directory.