Eddo's PhD repository

Content

In this repository I will share my PhD codes for supervised and non-supervised machine learning models for the quantification of lumbar paraspinal muscle health using conventional T₂-weighted MRI. The repository will contain programming codes (Python) for:

• Convolutional Neural Networks for the automatic segmentation of the lumbar paraspinal muscles

  Link to paper: https://www.nature.com/articles/s41598-022-16710-5

• Quantifying lumbar paraspinal intramuscular fat from clinical MRI

  Link to paper: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10869289/

Prerequisites

Make sure you have the following dependencies installed:

• Python 3.x
• NumPy
• Pandas
• SciPy
• scikit-learn
• nibabel

You can install all the dependencies by running:

pip install -r requirements.txt

This will install all the required packages listed in the requirements.txt file. Make sure you have pip installed and configured on your system.

Usage

To use the code, follow these steps:

1. Clone this repository to your local machine:

git clone https://github.com/Eddowesselink/PhD.git

2. Navigate to the code directory where you stored the repository

cd `/path/to/your/repository`

Thresholding

3. Run the script main_thresholding.py with the required arguments:

python main_thresholding.py --data_dir /path/to/your/data --kmeans --gmm

Replace /path/to/your/data with the path to the directory containing your MRI data. You can specify either --kmeans or --gmm to choose between KMeans or Gaussian Mixture Model clustering for segmentation.

CNN

3. Run the script main_CNN.py with the required arguments:

python main_thresholding.py --data_dir /path/to/your/data --model_dir /path/to/your/data 

Replace /path/to/your/data in -- data_dir with the path to the directory containing your MRI data. Replace /path/to/your/data in --model_dir with the path to the directory containing the model parameters.