Motion correction

During 2-photon calcium imaging, motion artifacts can pollute your raw recording and needs to corrected in order to enable analyses using ROIs (Regions Of Interest).

drim2p uses the SIMA (DOI) library to carry out motion correction.

Motion settings

Motion correction requires some additional information related to the strategy to use as well as the maximum expected displacement of pixels throughout the recording. This information is passed along to drim2p using a settings file. For this tutorial, we will use the default file available from the repository. Once you understand the workflow a bit better, you can customise it as described in the how-to guide.

In order to inform drim2p of which settings to use, you should download the default settings file and move it to the directory where your recording is.

You should then have the following file structure:

tutorial/
├── imaging_file.h5
├── imaging_file.ini
├── imaging_file.raw
├── imaging_file.xml
└── settings.toml.base  (NEW)

Motion correcting

From there, motion correction is as easy as conversion. All you need to do if run the following command:

drim2p motion correct . --settings-path settings.toml.base

If all goes well, you will see some output along these lines:

Applying motion correction for 'imaging_file' using DiscreteFourier2D.
Finished motion correction in 0h 1m 0.00s.
Saved motion correction to file.

You will now see two new files in your directory:

tutorial/
├── imaging_file.h5
├── imaging_file.ini
├── imaging_file.raw
├── imaging_file.xml
├── imaging_file_displacements.npz             (NEW)
├── imaging_file_motion_correction_report.txt  (NEW)
└── settings.toml.base

Displacements

The imaging_file_displacements.npz file is a compressed NumPy file containing information about the frame-to-frame displacements value used for the motion correction (under the 'displacements' key).

Report

The imaging_file_motion_correction_report.txt file is a plain text file containing basic information about what settings were used for the motion correction as well as information about how long it took and which file was corrected.

For our example imaging file, it looks like this:

Chunk: imaging_file             (name of the file processed)
Strategy: Fourier               (name of the strategy used)
Displacement: [50, 50]          (maximum X and Y displacements allowed)
Processing time: 0h 1m 0.00s    (time taken to motion correction)

What's next?

After motion correction, you will now be able to select ROIs on your recording in order to then analyse their signals. drim2p provides a GUI (Graphical User Interface) to draw ROIs on.