Deep Regression Segmentation for Cardiac Left-Ventricle MRI

This project was developed during the 2019 Neural Network Course held by Prof. Uncini at Sapienza University of Rome.

It is inspired on Deep Regression Segmentation for Cardiac Bi-Ventricle MR Images paper by X. Du et al.

Since the right ventricles dataset is not open source, we adapted the project only on left ventricles (which is open source and is available in this repository).

Summary

• Getting Started
• Key Points
• Authors
• License
• Acknowledgments

Getting Started

The project contains only a Jupyter Notebook file. Meet the prerequisite and use it. Google Colaboratory is recommended.

Prerequisites

You have two ways of meeting the prerequisites

• First (recommended, online)

  • Use Google Colaboratory
  • Done.

• Second (offline)

  • Python3
  • Jupyter pip install jupyterlab
  • Tensorflow (via conda or pip)
    • conda install tensorflow-gpu
    • pip install tensorflow-gpu
  • OpenCV pip install opencv-python
  • Matplotlib pip install matplotlib
  • Numpy pip install numpy
  • Scipy pip install scipy
  • Scikit-image pip install scikit-image
  • Scikit-learn pip install scikit-learn
  • Shapely pip install shapely
  • OSGEO pip install osgeo

Key Points

In this section, some recap images of the project are presented.

Preprocessing

Feature extraction

Coordinates extraction

Overview

Qualitative Analysis

Some qualitative examples are reported. For quantitative examples (Pearson's correlation coefficient, Dice Metric, Hausdorff distance) check the notebook.

Authors

• Giovanbattista Abbate - giabb
• Fabio Di Spazio - FabioDiS

License

This project is licensed under the MIT License - see the LICENSE.md file for details

Acknowledgments

• Billie Thompson - Provided README Template - PurpleBooth