A Novel Light U-Net Model for Left Ventricle Segmentation Using MRI

Published in Mathematics (MDPI), 2023

Authors:
Mehreen Irshad, Mussarat Yasmin, Muhammad Imran Sharif, Muhammad Rashid, Muhammad Irfan Sharif, Seifedine Kadry


This work proposes a lightweight U-Net architecture for accurate left ventricle segmentation from MRI images. The model is designed to reduce computational complexity while maintaining high segmentation performance, making it suitable for efficient medical image analysis.


Contributions 📃

  1. A lightweight U-Net model for medical image segmentation.
  2. Improved segmentation accuracy for left ventricle detection.
  3. Reduced computational complexity compared to standard U-Net.
  4. Evaluation on MRI datasets demonstrating robust performance.

📖 Citation (BibTeX)

@article{irshad2023novel,
  title     = {A Novel Light U-Net Model for Left Ventricle Segmentation Using MRI},
  author    = {Irshad, Mehreen and Yasmin, Mussarat and Sharif, Muhammad Imran and Rashid, Muhammad and Sharif, Muhammad Irfan and Kadry, Seifedine},
  journal   = {Mathematics},
  volume    = {11},
  number    = {14},
  pages     = {3245},
  year      = {2023},
  publisher = {MDPI},
  doi       = {10.3390/math11143245}
}

Paper PDF