DICOM4MICCAI Hands-on
  • Introduction
  • Prerequisites
    • If you are attending the tutorial in-person
    • If you are following on your own
  • Using DICOM to store your analysis results
    • Using 3D Slicer to convert non-DICOM segmentation results
    • DICOM Structured Reporting for radiomics
    • Using 3D Slicer for storing analysis results in DICOM
      • Step 0: 3D Slicer interface basics
      • Step 1: Import DICOM data
      • Step 2: Load DICOM image
      • Step 3: Segment lesions
      • Step 3.1:QuantitativeReporting interface overview
      • Step 3.2: Create and initialize a new segment
      • Step 3.3: Segment the lesions
      • Step 4: Explore and store the analysis results in DICOM
      • Step 5: Reload the analysis results from DICOM
      • Exporting DICOM data from 3D Slicer
    • Using MITK Workbench for storing segmentation results in DICOM
      • Step 0: MITK interface basics
      • Step 1: Import DICOM data
      • Step 2: Load DICOM image
      • Step 3: Segment lesions
      • Step 4: Explore and store the segmentations in DICOM
      • Step 5: Reload the segmentations from DICOM
  • DICOM data wrangling
  • Further reading
Powered by GitBook
On this page
  • Background
  • Hands-on exercise notes
  1. Using DICOM to store your analysis results

Using 3D Slicer to convert non-DICOM segmentation results

PreviousUsing DICOM to store your analysis resultsNextDICOM Structured Reporting for radiomics

Last updated 6 years ago

In this part of the tutorial we will start with a chest DICOM CT dataset, and a segmentation of a nodule stored using NRRD format. We will use this example to illustrate the differences and advantages of the DICOM Segmentation representation, and to demonstrate the steps of generating such standardized representation using 3D Slicer.

Background

DICOM provides various means to store segmentations, with the most commonly known being DICOM Radiotherapy (RT) Structure Sets (RTSS). RTSS represents segmentation as a set of planar contours. In this tutorial we will discuss the use of DICOM Segmentation image, which represents segmentation as a labeled image voxels. The differences between DICOM RTSS and DICOM Segmentation are discussed in .

DICOM Segmentation object is defined in of the DICOM standard . The picture below shows the high-level organization of a subset information that can be stored in this object.

Note the items that are marked with "CODE": those correspond to triplets of (CodeValue, CodingSchemeDesignator, CodeMeaning) attributes.

Hands-on exercise notes

  • Load LIDC dataset and corresponding segmentation (DICOM4MICCAI-2018/DICOM4MICCAI-Data-part1/LIDC*)

  • discuss the limitations of a representation

  • populate metadata using Quantitative Reporting module

  • export as DICOM

  • reload

  • export to file system

  • discuss the organization of the DICOM Segmentation

  • examine the contents of the resulting DICOM file (Atom editor)

Metadata for the individual segments is stored in the > . The content of those attributes can be explored with the .

batch conversion: ,

Segmentation Image Module
Segment Description Macro
dcmqi web app
dcmqi
dcmqi web app
this article
this section