Example Box: MRI artifacts


The aim of this example is to become familiar with different structural images and a range of artifacts.

This example is based on tools available in FSL, and the file names and instructions are specific to FSL. However, similar analyses can be performed using other neuroimaging software packages.

See the Getting Started page for instructions about downloading data and running FSL if you are new to either working with a terminal or with FSL.

Please download the dataset for this example here:

Data download

Once it is downloaded, move it to your preferred working directory and unzip it there.

This dataset contains a mixed set of images, including a range of structural modalities and examples both with and without artifacts.

Viewing the First Image

There is a large set of images here, consisting of different modalities and a mixture of images with and without artifacts. The objective here is to get used to viewing images and be able to tell if there are serious artifacts present or not. To start with we will view an image without artifacts: STRUCT0001.nii.gz

Load this image into a viewer (e.g. fsleyes) and inspect it. The image is a FLAIR image (dark CSF plus bright gray matter) and contains a small amount of bias field (certain areas and lighter or darker than others) but nothing that would be problematic for any analyses. Try adjusting the brightness and contract sliders to see the difference in tissues and structures more clearly, especially the deep grey matter structures.

Viewing the Other Images

Now go through and look at each of the other images in a viewer. We recommend using a separate viewing window for each as otherwise the viewer might struggle with the amount of data. Once you've examined an image feel free to close that viewer. In each case look at the image and try to find any artifacts, noting that there will be different contrasts (e.g., T2-weighted, FLAIR, etc.) and different resolutions (e.g., some with quite thick slices). Hence in some cases it will be more difficult to see anatomical structures clearly, but that would not be considered an artifact. The artifacts present in this data cover most of the ones discussed illustrated in the Primer.

For information about the different images click on the name below, but have a look at the image first and make your own mind up before checking here.

  • STRUCT0002.nii.gz

  • STRUCT0004.nii.gz

  • STRUCT0006.nii.gz

  • STRUCT0008.nii.gz

  • STRUCT0010.nii.gz

  • STRUCT0012.nii.gz

  • STRUCT0014.nii.gz

  • STRUCT0016.nii.gz

  • STRUCT0018.nii.gz

  • STRUCT0020.nii.gz

  • STRUCT0022.nii.gz

  • STRUCT0024.nii.gz

  • STRUCT0003.nii.gz

  • STRUCT0005.nii.gz

  • STRUCT0007.nii.gz

  • STRUCT0009.nii.gz

  • STRUCT0011.nii.gz

  • STRUCT0013.nii.gz

  • STRUCT0015.nii.gz

  • STRUCT0017.nii.gz

  • STRUCT0019.nii.gz

  • STRUCT0021.nii.gz

  • STRUCT0023.nii.gz

  • STRUCT0025.nii.gz

This example should give you some practical experience of what different modalities look like and how to identify a range of artifacts. In practice, when you are starting off and you think there is an artifact in an image then you should bring this to the attention of someone more experienced, who can advise you on the best course of action.