2014 Project Week:Quantitative Index Computation

• 

  Projects List

Key Investigators

• University of Iowa: Reinhard Beichel, Ethan Ulrich
• BWH: Andrey Fedorov, Nicole Aucoin
• GE: Jim Miller

Project Description

Objective

• We are developing tools for quantitative image analysis for PET scans, so far specifically involving head and neck cancer. The goal is to have a Slicer extension that supports the calculation and storage of quantitative indices from regions of interest. These indices go beyond those in the Label Statistics module, and can be computationally expensive.

Approach, Plan

• Review the desired functionality of the module and how it may be implemented in a workflow.
• Discuss appropriate type of module (CLI, scripted, loadable) if necessary.
• Discuss how calculated indices may be stored in Slicer environment.

Progress

• A simple CLI module has been started that calculates a few indices, but this does not allow for storage of the calculated values.
  • Check the SlicerExecutionModel CLI for an example of returning a double array, using the measurement tag, output channel
• The module currently can calculate mean, RMS, 1st quartile, median, third quartile, upper adjacent value, SAM, and peak. The calculated values are stored in the MRML scene and may be accessed from the vtkMRMLCommandLineModuleNode using the GetParameterDefault() method.
• The module may be downloaded as an extension from this repository.