added comment regarding bad flattening algorithm

canlab · Jan 9, 2024 · f2c5dd5 · f2c5dd5
1 parent 2372c6e
commit f2c5dd5
Showing 1 changed file with 6 additions and 0 deletions.
diff --git a/CanlabCore/@atlas/select_atlas_subset.m b/CanlabCore/@atlas/select_atlas_subset.m
@@ -229,6 +229,12 @@
     % -------------------------------------------------------------------
     obj_subset.dat = single(obj_subset.dat > 0);
 
+    % this is wrong. Consider, you could have a voxel that has 50% 
+    % probability of being voxel A and 50% probability of being voxel B.
+    % the max value is %50, but the probability that it's one or the other
+    % must be 100%. Let's try assuming conditional independence instead.
+    % There are a number of assumptions we need to make for a better
+    % solution though, so let this be the default
     obj_subset.probability_maps = max(obj_subset.probability_maps, [], 2);
 
     % Add labels for combined mask