Extraction of the blood vessel system of a liver is a challenging task in the field of medical image processing. Normally, doctors must examine each slice manually to achieve accurate vessel segmentation. Our solution is called Extraction of Liver Vessel Systems From CT-Image, an automatic vessel segmentation.
This is also a scientific research I and Duong, a friend of mine, conducted together. Our proposed model is called: MultiResDenseUNet. Here is the link to our paper Extraction of Liver Vessel Systems From CT-Image.
This project is trained and evaluated on the public IRCAD dataset.
From left to right: Original Input, Groundtruth, Prediction, and Intersection (Red is Difference and Green is Similarity)
My application is built with these frameworks and tools:
To get started, you should have prior knowledge on Python and Tensorflow at first. A few resources to get you started if this is your first Python or Tensorflow project:
-
Clone the repo
git clone https://github.com/phkhanhtrinh23/vessel_segmentation.git
-
Use any code editor to open the folder vessel_segmentation.
-
Run
preprocess_train_data.pyandpreprocess_valid_data.pyto preprocess and generate train and validation CT-Images, and then save them in the folder./data/artery. -
Open the notebook
train.ipynband follow the pipeline in that notebook. -
The folder
resultsis the folder of pre-saved weights and predictions that I have trained and validated. Up till now, our best Dice Score is 79.4%. -
Enjoy!
Contributions are what make GitHub such an amazing place to be learn, inspire, and create. Any contributions you make are greatly appreciated.
- Fork the project
- Create your Contribute branch:
git checkout -b contribute/Contribute - Commit your changes:
git commit -m 'add your messages' - Push to the branch:
git push origin contribute/Contribute - Open a pull request
Email: phkhanhtrinh23@gmail.com
LinkedIn: https://www.linkedin.com/in/trinh-pham-3103081a0/
Project Link: https://github.com/phkhanhtrinh23/vessel_segmentation.git