Skip to content

Visual inspection of the data

Checking data structure

Let’s take a look at the structure of the data you downloaded. The data is organised following BIDS - a standardised format for organising and naming neuroimaging files, and looks like this:

MoAEpilot
├── CHANGES
├── dataset_description.json
├── README
├── sub-01
│   ├── anat
│   │   └── sub-01_T1w.nii
│   └── func
│       ├── sub-01_task-auditory_bold.nii
│       └── sub-01_task-auditory_events.tsv
└── task-auditory_bold.json

Inside the folder containing the Mother of All Experiments (MoAE) data, you can find CHANGES, dataset_description.json, README and task-auditory_bold.json files - these files contain meta-data about the study. The subject directory, sub-01, contains anat and func folders, which store anatomical and functional images, respectively. Additionally, onset timings for task data are stored in the func folder alongside task images.

Inspecting anatomical data

An important step of any imaging analysis is to visually check your data for any problems that may be related to scanner issues, poor contrast, incidental findings, etc.

We will first look at the anatomical image sub-01_T1w.nii stored in sub-01/anat.

From your terminal or MATLAB window open SPM by typing:

spm fmri

or:

spm

and selecting fMRI in the SPM window.

To view an image, click on Display from the SPM menu and select the image you want to inspect from the pop-up window. The selected image will now be shown in the SPM graphics window.

File extensions

SPM can read uncompressed image files in NIfTI (.nii) and Analyze (.hdr/.img) formats. If your files are compressed with gzip and end in .gz, make sure to uncompress them first, e.g.: 

gunzip your_file.nii.gz

Browse through the image using your mouse or arrow keys. The crosshair will help you know what part of the image you are looking at. You can also right-click your mouse on the image to select the Movie feature and SPM will scroll through the image for you in x, y or z plane.

What to look for in anatomical data?

A few things that are worth paying particular attention to when reviewing anatomical images:

  • overall quality of the image (good contrast, no ring-like artefacts which can indicate excessive motion or image reconstruction issues),

  • correct image orientation,

  • abnormalities in the tissues (often showing as bright or dark spots where there shouldn’t be any).

Although initially, checking your data visually might seem daunting, you will notice that with practice you will quickly pick up this skill. mrishark.com has a rich repository of MRI data with various scanner-related artefacts and pathologies which can be a helpful resource when learning to distinguish between good and poor quality data.

Video walk-through

Inspecting functional data

Similarly to how we checked the anatomical data, we will now inspect a functional image.

Select Check Reg from the SPM menu and choose the functional data you want to view - in this case sub-01_task-auditory_bold.nii stored in sub-01/func. Browse through the image with your mouse or arrow keys just like you did with the anatomical data.

Loading 4D datasets into SPM

When loading the functional data into SPM, you may have noticed that the file name was followed by ,1. This indicates volume number in the time series - in this case volume one. You can change what volumes are being displayed in the SPM selection window by using the Filter option in the right bottom corner.

You can filter by:

  • specific volume, e.g. 1

  • several volumes, e.g. 1, 10, 100

  • range of volumes, e.g. 1:1000

  • all available volumes or a 4D file, e.g. Inf or NaN

Up to this point you have only loaded the first volume from your functional time series. To load the full time series, right-click on the SPM graphics window and navigate to Browse. In the pop-up window type Inf in the box below the Filter button and hit on your keyboard. This will prompt SPM to show all 84 volumes present in this directory. Right-click on the list of files and choose Select all. Press Done.

You will now see appear in the bottom-right corner of the SPM graphics window. Press the button to watch your functional time series as a movie.

What to look for in functional data?

Checking functional data is similar to checking anatomical data. You want to pay attention to any unwanted distortions, dark/bright spots which can indicate pathologies, incorrect orientation. Keep in mind that the spatial resolution of functional data is lower than that of anatomical data, so some of those issues may be more difficult to spot. Also, it’s worth noting that some distortions may be unavoidable in functional data. In the example dataset, you can see that there is some signal dropout towards the frontal pole and orbitofrontal regions. This is a consequence of the sequence used to acquire the data and cannot be fixed with preprocessing but is perfectly normal for this type of acquisition.

Another important part of viewing functional data is to visually check for excessive motion. You can do this by viewing the functional data as a movie, where each volume represents a timepoint in the scan.

Top tip

It’s a good practice to keep a spreadsheet where you note down your qualitative observations of the data at each preprocessing step. This can be useful down the line when trying to troubleshoot potential analysis problems.

Video walk-through