Heart ventricle segmentation

Left and right ventricle segmentation of Cardiac MRI

This repository shows the performance of a Pretrained U-NET applied to cardiac MRI slices.

1. Colab version in Colab file
2. Python version built in py file

The dataset used can be found in the file: MRI_dataset.rar. It cosists of MRI images with left and right ventricle annotations each.

This repository consists of the code of the preliminary version of the published articles:

1. Erik Carbajal-Degante, Steve Avendaño, Leonardo Ledesma, Jimena Olveres, Enrique Vallejo, Boris Escalante-Ramirez, "A multiphase texture-based model of active contours assisted by a convolutional neural network for automatic CT and MRI heart ventricle segmentation", Computer Methods and Programs in Biomedicine, Volume 211, 2021

2. Erik Carbajal-Degante, Steve Avendaño, Leonardo Ledesma, Jimena Olveres, Boris Escalante-Ramírez, "Active contours for multi-region segmentation with a convolutional neural network initialization," Proc. SPIE 11353, Optics, Photonics and Digital Technologies for Imaging Applications VI, 1135307 (1 April 2020)