Rotation plane misalignment

[lauraporta]

Is your feature request related to a problem? Please describe.
Even though the derotation works in the simulated data (even with different centres of rotation (see #24), in real datasets I am still observing wobbling and poor ROI alignment.

This could be due to the ellipsoidal shape of the rotation, which is likely due to the misalignment of the rotation plane and sampling plane. In such a case what we observe is an ellipsoidal projection in the imaging plane of a circular rotation happening in the rotation plane.

This consideration has also important implications regarding neuropil subtraction #17, as at different rotation angles we will be sampling a different slice of the cell, and therefore this could require the fit of a different neuropil coeficient for each angle.

Describe the solution you'd like
Therefore, I wonder if by modelling the rotation in a 3D stack (using the ellipse to infer the properties of it), if I can improve the quality of the derotation.

Describe alternatives you've considered
Expect the planes to be perfectly aligned with changes in the experimental setup.

[lauraporta]

According to Voigts 2020:

If the illumination of the FOV is not completely even, due to small deviations of the laser alignment from the objective or rotation axes or from non-uniform back-aperture illumination of the objective, rotation of the animal will bring imaged structures in and out of areas of higher or lower brightness.

As I understand it, it might not be fully solved also with an adjusted z-plane of rotation.

How they solved it:

$F_0$ is computed not over time, but across angles by computing a quantile for each angle first.

And for the neuropil subtraction:

We then applied neuropil correction by estimating the mixing of a background ROI devoid of any Ca 2+ transients on each ROI with a linear fit to the lowest 10 th percentile, and subtracting the neuropil data with this coefficient.

Further controls:

As an additional control, we removed ROIs where no Ca 2+ transients were evident over large > 60degree ranges of rotation for entire imaging sessions, as those ROIs could possibly have originated from cell bodies whose visibility might have been occluded in an angle-dependent manner by superficial blood vessels