search

AllenInstitute / brain_observatory_qc

Other
2 stars 0 forks source link

Inaccurate depths assigned to mesoscope containers #7

Closed DowntonCrabby closed 1 year ago

DowntonCrabby commented 3 years ago

Describe the bug Mesoscope experiments have 8 distinct imaging planes. Therefor, every mouse added to a mesoscope datacube generates 8 distinct containers.

The way mouse-seeks currently works, the value entered into the mongodb under db.cube_records.nominal_depth is iterated over the values entered at db.cube_records.area. Effectively, this means that if mouse is added to the datacube as Area: VISp_VISl_VISal_VISam Nominal_depth: 175_275 The following 8 containers would automatically generate: VISp_175 VISp_275 VISl_175 VISl_275 (etc)

This is a problem because occasionally, different visual area's are imaged at different depths. If I image VISp at 175 and 275, but I need to image VISl at 125 and 225, I have no way to control that.

I think that for container generation, we should not automatically iterate through area/nominal_depth values, but be able to indicate what depth all 8 individual planes should be.

Expected behavior I would expect to be able to individually set the area and depth for each FOV/individual experiment within a container so as to accurately reflect the area and depth.

Scope this is for all mesoscope sessions

Additional context

DowntonCrabby commented 3 years ago

this was originally a trello issue: https://trello.com/c/UZViuA4i/2-mesoscope-container-generation

matchings commented 3 years ago

There is also the issue that experiments in a container can have different depths from day to day depending on how the experiment was collected. I think it would make sense for multiple reasons to track both the actual imaging depth and a target imaging depth. The target depth can be used in container creation and would be a single value for all experiments in the container, and the actual depth can vary from day to day.

It will also be important to be able to set the target depths independently for each imaging plane, rather than iterating through set area / depth combinations.

DowntonCrabby commented 2 years ago

11/3/21 stakeholder group update: need information from @seanmcculloch and @matchings to get update and specific question for stakeholders group to answer

DowntonCrabby commented 1 year ago

Closing this issue