Edit this page

NA-MIC Project Weeks

Back to Projects List

Chest Imaging Platform Deep Learning Lung Segmentation

Key Investigators

Project Description

Integrate a lung segmentation algorithm based on Deep Learning (Keras+Tensorflow) into the Chest Imaging Platform. The goal is to make available in Slicer this and other similar tools based on Deep Learning.


  1. Integrate a Lung Segmentation algorithm based on Deep Learning in the Chest Imaging Platform.
  2. Make available these and other similar tools in Slicer

Approach and Plan

  1. Integrate a local python customized distribution in Slicer
  2. Run a full deep learning keras-tensorboard based pipeline

Progress and Next Steps

  1. The core infrastructure was integrated in the CIP library
  2. Full pipeline of lung segmentation was run successfully in Slicer
  3. The model needs more work

Background and References