Timeline for Listing applications of the SVD
Current License: CC BY-SA 4.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 17, 2021 at 22:23 | comment | added | Ander Biguri | Other more advanced methods use this signal to create motion models etc, but that goes a bit beyond the scope. The main goal of this is to be able to know in which stage of the respiratory cycle any given "frame" is at. | |
| Nov 17, 2021 at 22:22 | comment | added | Ander Biguri | @ogogmad generally you take the main 3 or 4 principal components and try to find which one creates singular values that fluctuate at breathing phase when applied to the data. Often is not the first one, but second or third PC. That will give you a "respiratory surrogate signal". Then what you do with it changes depend sin which motion compensation method you use. The "easy" one is to bin the "frames" into an amplitude (or singular value "value"), so you can reconstruct several images instead of one. | |
| Nov 17, 2021 at 21:06 | comment | added | wlad | Thank you for this example. I have some questions: Is this principle component usually the dominant one? Does the projection of a "frame" onto this principle component give you the phase? Do you subtract away the part of each frame parallel to this principle component to eliminate the effect that breathing has on each frame? I tried reading the paper, but I couldn't fully understand the model they used. | |
| S Nov 17, 2021 at 11:56 | review | First answers | |||
| Nov 17, 2021 at 12:53 | |||||
| S Nov 17, 2021 at 11:56 | history | answered | Ander Biguri | CC BY-SA 4.0 | |
| S Nov 17, 2021 at 11:56 | history | made wiki | Post Made Community Wiki by Ander Biguri |