search

Donders-Institute / PRESTUS

PREprocessing & Simulations for Transcranial Ultrasound Stimulation package
GNU General Public License v3.0
12 stars 10 forks source link

Fix heating plots for 'create_group_plots_MNI' #9

Closed KTZ228 closed 1 year ago

mekman commented 1 year ago

@KTZ228 could you please provide more information on what needs fixing? e.g. What is the current result and how is it supposed to look like?

KTZ228 commented 1 year ago

So this is what the heating plots currently look like: all_maxT_MNI_target_left_amygdala_max_intensity

And this is what they should look like: all_final_isppa_MNI_target_left_amygdala_max_intensity

mekman commented 1 year ago

Cheers. To summarise, we want

  1. a different colormap
  2. the min/max colormap values of the background image need to be adjusted, so that the anatomy becomes visible
  3. the min/max colormap values of the heat simulation overlay need to be adjusted

Anything else? It looks like there is some masking going on, as the top image has some high-intensity streaks/lines that are not present at the bottom.

Did you simply fix these min/max colormap values by hand, or do you propose to choose them algorithmically (e.g. min/max = bottom/top percentile of distribution)?

achetverikov commented 1 year ago

I think the heating map artefacts at the borders that create problems

On Tue, 23 May 2023 at 16:20, Matthias Ekman @.***> wrote:

Cheers. To summarise, we want

  1. a different colormap
  2. the min/max colormap values of the background image need to be adjusted, so that the anatomy becomes visible
  3. the min/max colormap values of the heat simulation overlay need to be adjusted

Anything else? It looks like there is some masking going on, as the top image has some high-intensity streaks/lines that are not present at the bottom.

Did you simply fix these min/max colormap values by hand, or do you propose to choose them algorithmically (e.g. min/max = bottom/top percentile of distribution)?

— Reply to this email directly, view it on GitHub https://github.com/achetverikov/PRESTUS/issues/9#issuecomment-1559505561, or unsubscribe https://github.com/notifications/unsubscribe-auth/AALF3TUS2PTYHCK5CO3N2J3XHTBSNANCNFSM6AAAAAAXGTJI6M . You are receiving this because you are subscribed to this thread.Message ID: @.***>

--

Best regards, Andrey Chetverikov

KTZ228 commented 1 year ago

As it stands, the colormap values are calculated in the first pass of the function. Indeed, the background is hidden behing the baseline overlay of 37. Aside from that, the streaks at the top should not be there so I was wondering how they got there in the first place. I think the colormap itself it fine but if you have a better suggestion please let me know.

mekman commented 1 year ago

@KTZ228 Sorry, can you be more precise on what you actually want to be fixed with this issue: the plotting, or the heating map artefact?

the streaks at the top should not be there so I was wondering how they got there in the first place.

I don't follow. You made both these plots I assume. What changed between plot 1 and plot 2 that could make the artefact go away?

KTZ228 commented 1 year ago

Found a solution, will upload in next commit.