This lesson is still being designed and assembled (Pre-Alpha version)

sMRI Quality Control

Overview

Teaching: 20 min
Exercises: 10 min
Questions
  • How do we identify image preprocessing failures?

Objectives
  • Visualize processing failures

  • Familiarize with automatic QC tools

You Are Here!

course_flow

Things that can go wrong…

Acquistion

Due to MR physics (e.g. Field of view (FOV), ghosting, aliasing)

Incorrect parameters can truncate and/or duplicate brain anatomy.

Drawing

Due to participant (e.g. Motion artifacts)

Drawing

Quantification

Exisiting image processing pipelines (e.g. FreeSurfer, CIVET) will have a few QC tools and examples that can help with failure detection and quality control of volumetric segmentations and surface parcellations.

Drawing

Usage of new method will require your own QC protocols. Especially for highly specific segmentation methods require visual inspection from a neuroanatomy expert. Even for the qualitiative visual inspection, it is important create a QC protocol and share it with the results.

HC_and_CB_MAGeT

Note: see Hippocampal and cerebellar for segmentation method details.

Automatic QC tools

Using reports from exisiting pipelines: https://fmriprep.org/en/stable/_static/sample_report.html

Drawing

Using QC tools

MRIQC: extracts no-reference IQMs (image quality metrics) from structural (T1w and T2w) and functional MRI (magnetic resonance imaging) data. (Developed by the Poldrack Lab at Stanford University for use at the Center for Reproducible Neuroscience (CRN), as well as for open-source software distribution.)

Individual report Group report
Drawing Drawing

VisualQC: assistive tool to improve the quality control workflow of neuroimaging data (Author: Pradeep Reddy Raamana).

T1w acquisition Alignment Cortical Parcellation
t1_mri_visual_QC alignment_mismatched_colormix_visualQC cortical_zoomed_in

Key Points

  • Image processing failures happen! It is important to perform systematic quality control to minimize biases