MEDimage-app-docs
  • 👋Welcome!
  • 👊Quick start
  • 👀Interface overview
  • 🧠Radiomics
    • Data Processing
      • Data Manager
      • Radiomics Pre-Checks
    • Feature Extraction
      • Batch Extraction
      • Single Scan
  • 👨‍đŸĢLearning
  • â™ģī¸Code generation
  • đŸ•šī¸Demonstration
  • 👩‍đŸ’ģContributing
  • 🤓About us
  • FORMS
    • âœī¸Contact us
    • âš ī¸Report an issue
  • MEDIA
    • đŸĨMEDimage website
    • 🤖GitHub
    • đŸ“ēYouTube
    • 👾Discord
Powered by GitBook
On this page
  1. Radiomics
  2. Feature Extraction

Single Scan

This is a tutorial on how to use the software to extract radiomic features from a single scan.

PreviousBatch ExtractionNextLearning

Last updated 11 months ago

A tutorial video is available at the bottom of the page.

Contrary to the Batch Extraction implementation, the single-scan feature extraction provides great flexibility through a drag-and-drop design. This enables users to customize their Radiomics analysis for a specific image. In this mode, users have control over the processing steps and the features to be extracted. Moreover, additional features such as 3D image visualization are available.

The following image summarizes the available resources in the single scan interface:

Having gained an understanding of the various components of the extraction interface, the following section illustrates the available nodes, highlighting their individual functionalities, inputs, and outputs.

Node
Description
Input
Output

None

Input node (With loaded file)

Imaging Volume & Mask

Imaging Volume & Mask

Interpolated Imaging Volume & Mask

Binary Mask

Segmented Binary Mask

Imaging Volume

Imaging Volume

Imaging Volume

Filtered Imaging Volume

Imaging Volume with ROI voxels only

Discretized Imaging Volume with ROI voxels only

Feature calculation is the final processing step. Upon clicking, a page will appear, allowing you to select the features for extraction.

The input depends on features type. Usually Imaging volume and ROI mask.

Features dictionary

To select the features to extract, click on the extraction node, a page will appear (see figure below) and you will be able to select feature families and features to extract for your analysis.

Tutorial video

The input node handles the loading of data into memory, supporting both DICOM files and objects. Additionally, it provides 3D visualization of the images.

binary object

The segmentation node is used to define the regions in which features are calculated. These regions are automatically identified from the loaded scan. More details in .

The interpolation node is used to interpolate to isotropic voxel spacing. Many features require interpolation prior to their extraction. More detail in .

The re-segmentation process is performed to exclude voxels from a previously segmented ROI, and is performed after interpolation. More details in .

ROI extraction node replaces voxels outside the ROI by a placeholder value, often NaN. More details in .

Filtering is conducted after image interpolation. A set of linear filters is available for use. More details in .

Discretization or quantization of image intensities inside the ROI is required for the calculation of texture features. More details in .

🧠
MEDscan
MEDscan
IBSI section 2.3
IBSI section 2.4
IBSI section 2.5
IBSI section 2.6
IBSI-2
IBSI section 2.7
Single image feature extraction interface
Page to select features to extract