2016 Winter Project Week/Projects/ChestImagingPlatformWorkflows

Key Investigators

• James Ross
• Raúl San José

Project Description

The Chest Imaging Platform (CIP) is a collection of C++ libraries, command-line executables, and python modules for segmenting, registering, processing, and quantitatively evaluating medical images of the chest; it is developed with large-scale, batch processing of high-resolution computed tomography (CT) images in mind. Many of the high-level workflows involve executing multiple CIP command-line tools and/or python modules, which imposes a barrier to entry for new users. Nipype is an open-soure, python-based software package that provides the ability to package multiple execution steps into a single workflow. In this project, I will be creating nipype-based workflows of commonly used, high-level tasks in order to make CIP functionality more accessible to new users.

Objective	Approach and Plan	Progress and Next Steps
Implement Nipype workflows for commonly used chest image analysis tasks	Focus on automatic lobe segmentation pipeline	Workflows exist for particle deployment and parenchyma phenotype computation End-to-end workflow created for automatic lobe segmentation A few extra nodes to add Include in next CIP release