visualqc 0.6.3 documentation

Gallery - Defacing algorithm accuracy

Gallery - Defacing algorithm accuracy
Gallery - Registration : comparison of spatial alignment
Gallery - Segmentation/ROI - anatomical accuracy evaluation 

Table Of Contents

  • VisualQC
    • Use-cases supported
    • Features
    • Manual
    • Galleries
    • Contributions
    • Citations
  • Installation
    • Stable release
    • Requirements
    • From sources
  • Use-cases supported
    • Structural MRI use-cases
    • Functional MRI use-cases
    • Diffusion MRI use-cases
    • Registration/Alignment use-cases
  • Recommended Usage
    • General suggestions:
    • For Freesurfer outputs:
    • Alignment checks (Registration quality)
  • Command line usage
  • Data formats and requirements
  • Review interface
    • Common elements
    • Review interface - Freesurfer
    • Review interface - Anatomical / T1w MRI
    • Review interface - Registration
    • Review interface - Functional MRI
    • Review interface - Diffusion MRI
    • Review interface - Defacing
  • Gallery - Freesurfer Parcellation
  • Gallery - Freesurfer (labels filled)
  • Gallery - Functional MRI scan - artefact detection and rating
  • Gallery - Diffusion MRI artefact detection and rating
  • Gallery - Registration : comparison of spatial alignment
  • Gallery - Defacing algorithm accuracy
  • Gallery - Segmentation/ROI - anatomical accuracy evaluation
  • Gallery - Structural T1w MRI - artefact detection and rating
  • Command line usage - Freesurfer
    • Input and output
    • Overlay options
    • Outlier detection
    • Layout options
    • Workflow
  • Command line usage - Defacing
    • Input and output
  • Command line usage - Functional MRI
    • Input and output
    • Preprocessing
    • Outlier detection
    • Layout options
    • Workflow
  • Command line usage - Diffusion MRI
    • Input and output
    • Preprocessing
    • Visualization
    • Outlier detection
    • Layout options
    • Workflow
  • Command line usage - Alignment
    • Input and output
    • Visualization
    • Outlier detection
    • Layout options
    • Workflow
  • Command line usage - T1w MRI
    • Input and output
    • Visualization
    • Outlier detection
    • Layout options
    • Workflow
  • Generating Screenshots
  • Example usage - Freesurfer
    • Generating and Downloading Required Files from a remote SSH server
  • Example usage - Segmentation
  • Example usage - Defacing
  • Example usage - Functional MRI
  • Example usage - Diffusion MRI
  • Example usage - Alignment
  • Example usage - T1w MRI
  • Contributing
    • Types of Contributions
      • Report Bugs
      • Fix Bugs
      • Implement Features
      • Write Documentation
      • Submit Feedback
    • Get Started!
    • Pull Request Guidelines
    • Tips
  • Citation
  • Credits
    • Development Lead
    • Contributors
  • VisualQC
    • Use-cases supported
    • Features
    • Manual
    • Galleries
    • Contributions
    • Citations
  • Installation
    • Stable release
    • Requirements
    • From sources
  • Use-cases supported
    • Structural MRI use-cases
    • Functional MRI use-cases
    • Diffusion MRI use-cases
    • Registration/Alignment use-cases
  • Recommended Usage
    • General suggestions:
    • For Freesurfer outputs:
    • Alignment checks (Registration quality)
  • Command line usage
  • Data formats and requirements
  • Review interface
    • Common elements
    • Review interface - Freesurfer
    • Review interface - Anatomical / T1w MRI
    • Review interface - Registration
    • Review interface - Functional MRI
    • Review interface - Diffusion MRI
    • Review interface - Defacing
  • Gallery - Freesurfer Parcellation
  • Gallery - Freesurfer (labels filled)
  • Gallery - Functional MRI scan - artefact detection and rating
  • Gallery - Diffusion MRI artefact detection and rating
  • Gallery - Registration : comparison of spatial alignment
  • Gallery - Defacing algorithm accuracy
  • Gallery - Segmentation/ROI - anatomical accuracy evaluation
  • Gallery - Structural T1w MRI - artefact detection and rating
  • Command line usage - Freesurfer
    • Input and output
    • Overlay options
    • Outlier detection
    • Layout options
    • Workflow
  • Command line usage - Defacing
    • Input and output
  • Command line usage - Functional MRI
    • Input and output
    • Preprocessing
    • Outlier detection
    • Layout options
    • Workflow
  • Command line usage - Diffusion MRI
    • Input and output
    • Preprocessing
    • Visualization
    • Outlier detection
    • Layout options
    • Workflow
  • Command line usage - Alignment
    • Input and output
    • Visualization
    • Outlier detection
    • Layout options
    • Workflow
  • Command line usage - T1w MRI
    • Input and output
    • Visualization
    • Outlier detection
    • Layout options
    • Workflow
  • Generating Screenshots
  • Example usage - Freesurfer
    • Generating and Downloading Required Files from a remote SSH server
  • Example usage - Segmentation
  • Example usage - Defacing
  • Example usage - Functional MRI
  • Example usage - Diffusion MRI
  • Example usage - Alignment
  • Example usage - T1w MRI
  • Contributing
    • Types of Contributions
      • Report Bugs
      • Fix Bugs
      • Implement Features
      • Write Documentation
      • Submit Feedback
    • Get Started!
    • Pull Request Guidelines
    • Tips
  • Citation
  • Credits
    • Development Lead
    • Contributors

Gallery - Defacing algorithm accuracy¶

VisualQC’s Defacing module can help quickly evaluate the accuracy of defacing algorithms by presenting a comprehensive picture of the result of “defacing”, and whether it led to data loss, or it failed to remove potentially reidentifiable info (such as facial features). Various examples below illustrate the defacing QC interface along with different types of failures by the defacing algorithm. Please note 1) the overlay in red on top of green MRI helps to highlight what has been removed/stripped to achieve defacing, and 2) that we may have blurred and/or greyed out some areas to avoid revealing any info remotely helping with patient reidentification.

An example result that correctly defaced a T1w MR image without over- or under-stripping:

_images/defacing_illustration_1.png

Below we observe overstripping of the brain the frontal areas, which can be noticed both in the 3D render as well as in the composite overlay:

_images/defacing_illustration_overstrip_frontal.png

The example below shows the defacing algorithm understripping the brain wherein eyes and few other facial features haven’t been removed when they should have been. We greyed out some areas to protect the patient’s privacy as much as possible.

_images/defacing_illustration_cutbehindface.png

This example shows the algorithm going completely bonkers:

_images/defacing_illustration_maskmisalign.png
© Copyright 2017, Pradeep Reddy Raamana. Built with Sphinx 6.1.3 and the Sizzle theme.