Model Zoo Elastix

autoTICI: Automatic Brain Tissue Reperfusion Scoring on 2D DSA Images of Acute Ischemic Stroke Patients

Registration Description

"inter/intra-patient; affine transformation; mutual information"

Image data

Cerebral Digital Subtraction Angiography (DSA): 2D X-Ray image sequence with dimension: 1024 x 1024 x sequence length

Screen shot:

alt-text

Examples of DSA images.

Application

The registration was used to align frames of an acquired DSA sequence, align the minIP images before and after endovascular treatment (EVT) and align the minIP images across patients. With this registration, motion artifacts caused by patient motion could be mitigated. Intra- and inter- patient DSA could be registered, which is useful for mapping annotated regions on atlas/templates onto specific patients. This registration is part of a pipeline for automatical brain tissue reperfusion scoring algorithm.

Screen shot:

alt-text

An example of DSA sequence motion correction using simple elastix.

Registration Settings

For parameter files see the Elastix Model Zoo repository on GitHub.

Elastix version: v2.0rc1 (Simple Elastix)

Command line call: elastixImageFilter = sitk.ElastixImageFilter() elastixImageFilter.SetFixedImage(fixed_image) elastixImageFilter.SetMovingImage(moving_image) parameterMap = elatixImageFilter.ReadParameterFile('Par0062.txt') elastixImageFilter.SetParameterMap(parameterMap) elastixImageFilter.Execute() aligned_img = elastixImageFilter.GetResultImage()

Published in

Su et al. (2020) autoTICI: Automatic Brain Tissue Reperfusion Scoring on 2D DSA Images of Acute Ischemic Stroke Patients (submitted)