jmeehan16 / whitematter Goto Github PK

if time allows

actually the list.wsgi should not return the different arrays - but it will return the number of dimensions in a single array, which is the number of brain volumes. So it should do a query like
SELECT DISTINCT dimension FROM image WHERE {some rules so it is not searching ALL the records}

in that way If there are 16 volumes
it should return a json like
{"brainvols": ["0", "1","2","3","4","5","6","7","8","9","10","11","12","13","14","15"]}

also after this is done the relevant code in javascript and python should be changed so the name of the array (e.g "image") is hardcoded and not given by the frontend

In other words, remove the hardcoded slice depth in axis z (now 90) and have that number be sent by the frontend with a slider.

Now we use dropdown menus to select brain and which view to be used each time. These two dropdown menus and the show button should be substituted with a drag-n-drop mechanism.
The plan is to have a vertical scrolling pane in the left side having the brain labels and these labels are draggable. The rest of the window is droppable, so by dropping the label the viewer will be initialized only if this label is not there already.

things to do:

• Redesign frontend/backend
  • 3 views per brain volume
  • 4 brains displayed at once
  • menu for choosing brains
  • method of selecting slice
  • coordination of selected slice
• Import all brain volumes
  • write a script to iterate nifti file by nifti file, importing each into image array
  • imgtocsv code will have to be changed to support multiple volumes
  • implement indeces

1. Change import code to handle multiple brains into one array
2. Design API
3. Settle on a front-end design
4. Create panels for front-end
5. Change Python methods according to API
6. Populate 3 views for one brain volume
7. Develop coordinated views for multiple brains
8. Ability to choose from multiple brains

the idea is to put an one pixel border with a distinct color for each view so that when we slide e.g the top view - these colored lines move accordingly to the other views..

some brain volumes don't have color intensities between 0...255, so we have to normalize
better to do it on the fly after it is being stored in scidb
-edit that's not a good Idea because for each slice we have another scidb query and it will be very slow