20171207
overlapping structures at edges
This is a possible method for looking for structures that overlap and do not overlap. Test is with mitochondria and ERZ as labeled with mitotracker and ER tracker.
Processed shows green objects that do not overlap red (mito that do not overlap ER) in green, mito that do overlap ER in blue, and ER that does not overlap mito in red.
The problem is that there is a 5 to 7 seconds delay between the red and green image at each timpoint. Therefore, in this example the differences are probably exclusively due to movement. Either there is no overlap at any timepoint or the red needs to be bg subtracted. (Constant? Rolling ball?)
Imarus should have a touching objects funtion to, but taking it apart would probably reveal the same basic math.
selectWindow("shdia airytest 15secT_Airyscan Processing.tif");
run("Split Channels");
selectWindow("C2-shdia airytest 15secT_Airyscan Processing.tif");
run("Enhance Contrast...", "saturated=0 normalize process_all use");
selectWindow("C1-shdia airytest 15secT_Airyscan Processing.tif");
run("Enhance Contrast...", "saturated=0 normalize process_all use");
imageCalculator("Difference create stack", "C1-shdia airytest 15secT_Airyscan Processing.tif","C2-shdia airytest 15secT_Airyscan Processing.tif");
rename("difference");
selectWindow("C1-shdia airytest 15secT_Airyscan Processing.tif");
saveAs("Tiff", "R:\\MicroscopyCore\\mcammer\\Schmidt lab\\Devi Thiagarajan\\20171207_mitochondira_ER\\overlap math\\MITO-shdia airytest 15secT_Airyscan Processing.tif");
selectWindow("C2-shdia airytest 15secT_Airyscan Processing.tif");
saveAs("Tiff", "R:\\MicroscopyCore\\mcammer\\Schmidt lab\\Devi Thiagarajan\\20171207_mitochondira_ER\\overlap math\\ER-shdia airytest 15secT_Airyscan Processing.tif");
imageCalculator("Subtract create stack", "difference","MITO-shdia airytest 15secT_Airyscan Processing.tif");
selectWindow("Result of difference");
rename("difference minus mito");
imageCalculator("Subtract create stack", "difference","ER-shdia airytest 15secT_Airyscan Processing.tif");
selectWindow("Result of difference");
rename("difference minus ER");
run("Merge Channels...", "c1=[difference minus mito] c2=[difference minus ER] c3=[MITO-shdia airytest 15secT_Airyscan Processing.tif] create keep ignore");
This one shows the method works, but the reason the mitochondrion isn't overlapping with ER is because the mitochondria jumped faster than the two channel imaging. Or maybe it is real that the mitochondrion escaped the ER for up to 15 seconds. Who knows?
This version incorporates global constant bg subtractions based on the mean intensities of the nucleii. Method (with other spurious commands mixed in) here.
But this simulation says not really working as advertised, but we'll get the method to work...
