help > hmrMotionArtifactByChannel
Jun 29, 2018  07:06 PM | Jessica Defenderfer - University of Tennessee Health Science Center
Hello, looking for some advice from experienced and knowledgeable NIRS scientists - I am working through some data obtained from the temporal and frontal regions on the left hemisphere (thirteen 30mm channels, one 10mm channel). I am becoming more aware of the potential unwanted effects/artifacts from the temporal muscle. I have a few questions about processing these data:

1) Is hmrDeconvHRF_DriftSS the only option for regressing short channel data from long channels? There seem to be a lot of other parameters built in to that one step that I don't think I need (i.e. convolution, temporal basis functions). 

2) Since I feel that motion artifact may not be common across all channels in my probe design, I decided to try hmrMotionArtifactByChannel. From what I've read, this option will essentially do the same thing as hmrMotionArtifact, but identify artifact specific to each channel (not marking it on all other channels). So far, it doesn't appear to be working this way. For example, one participant has several ampthreshold violations on 2 channels, but in their procResults, none of the trials were accepted across all channels and conditions. I may be misunderstanding what I should expect, but my initial expectation was to see the number of accepted trials to be varied across channels and conditions. 

3) I only have one short distance probe. Would it be effective to also run PCA in addition to regressing out the short probe? I realize that the scalp fluctuations may be different on other areas of the scalp.