Map quality

tonydoss

I've produce an image of a region of a rat brain, with 20 um spot size, 400 um/sec scan speed. Each line has an individual ICP data file, matched to the laser log.
Somehow, it looks a bit "dribbly": !(src)[https://photos.app.goo.gl/HQd5YkNi1SX2Yqwp8]

then, using the same settings, I've tried by merging all the lines into one (so there's one ICP data file for the whole map). This looks a bit better:
(this is the region next to the one in the map above - the zoning is just because I had split the analysis in two so I could have a standard after 30', and I didn't use an internal standard to account for the drift)
but of course, it's a pain because I can't use the laser log so I have to define selections for the sample using channels.
The upside is that it takes a third of the time to produce the same area (1 hr; versus 3 hr above). That's because is the experiment above I have 10 sec laser warm up before and 10 sec wash out after each line.

Any chance you could provide advice on how to get a better map? (I did watch the webinar about mapping, which was brilliant!)

thanks,
Anthony

Bence

How have you aligned your scanlines in the image setup panel? I.e. have you used 'Start', 'End', 'Centre', or 'Justified' in the option below Cache and above Map Style in the top left?

The reason that the image looks a little unaligned is that each scan line is slightly different in length (presumably because you're creating selections from channel values) and that they are not aligned.

My advice would be to nail down the start of each scan line with multiple criteria (e.g. Cu_ppm > x, Zn > y etc etc). Then, with nice and sharp starts to each scanline, I would align them to using the 'Start' option.

You'll notice above that the criteria I've suggested are the ppm channels. Remember that you can calculate ppm values without any sample selections at all, and then go back and select your samples. The advantage of using ppm channels is that they have been corrected for baselines and sensitivity drift. At a pinch you can try the baseline-subtracted _CPS channels; they will at least have baselines subtracted). This should give you a more precise threshold for setting the start of each selection (the end is not as important because you're not aligning your scanlines using their end, and any additional data selected will just appear on the right side of your image, if you're scanning left to right).

Please let me know if any of that doesn't make sense, or improve your images.

tonydoss

Bence I've used 'Start', but either options (even 'Justified') don't seem to make much of a difference

Again, this is what I get when there's one ICP data file per line and I use the import info to create the selections (I found some issues with the laser log... that's for another post): map
it turns out the lines still don't have the same length because they have slightly different durations (beats me to figure out why since they have the same pattern length on the laser and scan speed; maybe something to do with the iCAP?) and they are all cropped by the same amount (15 sec from start; 9 sec from end).

Merging all lines into one seems to give an image a bit less noisier: map (that's the region adjacent to the map above; don't mind the zoning), and like I said it's 3 times faster (because no laser warm up nor wash out for each line)... but I'd rather have one ICP data file per line as you suggest in the webinar.

tonydoss

but the (more) dribbly image is the one where I didn't use channel values, but where I have an ICP data file (and corresponding laser log) for each line: map
why aren't lines aligned if I have an ICP data file for each line? (I might be missing something here... mmm long day haha)

Bence

tonydoss It definitely seems like timing issues. You can see from the maps that the scan lines are different lengths and that the maps starts in different places.

tonydoss I'd rather have one ICP data file per line as you suggest in the webinar.

I'm ambivilent about this actually 😄 Whichever works best is my suggested approach.

If you want to send us a copy of the session to the support email address, we can take a look to see what can be done to improve the alignment, but I think longer term, those timing issues will need to be sorted for best results.