Back to PLS HelpPosted on 11/06/20 15:25:27
Number of posts: 8
Hello everyone,
My lab is looking at analyzing a dataset of T1 images that was collected on several different scanners. In a previous post titled “Structural PLS”, it was mentioned that PLS may not be suitable to use for analysis if there are differences in resolution, image intensity, contrast, field strength, etc. The dataset that we are looking to investigate offers data that is preprocessed to account for differences in resolution, intensity (e.g. are normalized), have been B1 field-corrected, and have the same field strength, despite being collected at different sites. Would a PLS analysis of these data be interpretable if the corrected data were used? Would it help at all to include collection site as a covariate in this case?
Thank you,
Alex
Replies:
Posted on 11/06/20 15:29:27
Number of posts: 394
Hello everyone,
My lab is looking at analyzing a dataset of T1 images that was collected on several different scanners. In a previous post titled “Structural PLS”, it was mentioned that PLS may not be suitable to use for analysis if there are differences in resolution, image intensity, contrast, field strength, etc. The dataset that we are looking to investigate offers data that is preprocessed to account for differences in resolution, intensity (e.g. are normalized), have been B1 field-corrected, and have the same field strength, despite being collected at different sites. Would a PLS analysis of these data be interpretable if the corrected data were used? Would it help at all to include collection site as a covariate in this case?
Thank you,
Alex
Hi Alex - the scanne/site differences would be a problem for any analysis, not just PLS. Having said that, if you processing steps correct for it, you should give it a try and definietly include "site" as a factor (covariate) to see if it plays out still.
I'm Online
Posted on 11/06/20 15:34:38
Number of posts: 229
However, do keep in mind that the values in your T1w images mean nothing - so you should not submit T1w images to PLS (or any analysis).
You can submit GM probablility maps to PLS, with the caveat that the preprocessing pipelines will have different effects on the GMprob calculations.
If you have quantitative T1 measures, the scanner differences may be less problematic, but it doesn't sound like that is what you have.
cheers
Nancy
Login to reply to this topic.
