Example Dataset for Knogler et al., 2018
Here, a functional imaging dataset is provided, as shown in Figure 1 and Figure 2.
It is a compressed HDF5 file, that has the following structure:
- anatomy (Dataset, 100 planes, 320x334 px)
- anatomy_mask (Dataset, 100 planes, 320x334 px)
- behavior (Group, List (indexed by planes), pandas DataFrame)
- imaging_data (Dataset, 100 planes, 860 frames, 320x334 px, pre-processed)
- regressors (Group, List (indexed by planes), dictionary)
- stimuli (Pandas DataFrame)
The behavior group offers the data acquired during imaging.
- the timestamp
- the tail
- the vigor (standard deviation of a 50 ms rolling buffer)
- the grating speed (OMR, mm/s, OKR, rotation angle)
- swimming (binary, if fish is swimming)
- stimulus id (see stimuli group in HDF5 file)
- stimulus associated parameter (see stimuli group, e.g. speed in mm/s for OMR)
- raw left eye trace (degrees)
- raw right eye trace (degrees)
Here, the regressors built from sensory (see stimuli group) or behavioral (see behavior) parameters are provided.
They are already convolved with the calcium kernel and interpolated to fit the timesteps of the imaging.
Imaging data was pre-processed:
- Registered in 2D
- Registered in 3D
Pandas DataFrame providing the complete stimulus presented.
Sum of _imagingdata across time.
anatomy_mask is based on the anatomy and a manual selected mask of the foreground, i.e. Purkinje cells.