Please enable JavaScript.
Coggle requires JavaScript to display documents.
fNIRS (fNIRS study,
--> compared measurements of "yes"…
fNIRS
fNIRS study,
--> compared measurements of "yes" answers to subjects intended answer :3 !! ---> thats how accuracy was decoded #-
-
-
-
difference between MCP etc :P
---> uses all electrodes
--> 11 people succeeded / above 70% success level (vs 9 for GLM)!!
-
-
-
Experiment: Block design
--> 20, 10 seconds mental drawing per run +
--> alternated with 21, 10 sec baseline do nothing blocks
------> 20 seconds per run = all in all 10 minutes :3 !!
--> imagine drawing a simple geometrical shape or outline :3 !!
Equipment:
--> 3 source + 6 detector opcodes
--> 10/20 placement system
-----> above left front parietal sensorimotor cortex :3!!
ANALYSIS:
GLM
--> first best channel was selected (Localizer)
-----> by GLM analysis for deox and oxygenated blood seperately (by contrast mental droawing with baseline)
--------> then based on criterion (by averaging oxygenated blood and deoxygenated blood per channel) <-- channel with highest blues for both during motor task = selected !!
Cut off value = mean of all yes and no conditions (criterion) applied to rest of 4 trial runs
---> then localizer checked again too see if its still same :3 !
Brain computer interfaces( BCI's)
--> cool for ppl in vegetative states and paralyzed people and ppl with depression, pain etc (neurofeeback :D!!) :3!!
--> basically output given by brain, computer does something, computer output = used as feedback, and changes are made to lead to better output :3!! Eg. for neurofeedback display of brain activity, for wheelchair user actual movement of wheelchair :3 !!
-
-
what is it ?
--> software / hardware of relating brain activity to computer signals without involvement of motor neurons etc
--> directly connection to the brain basically, bypasses periphery neurons/nerves and muscles
how does it work
--> machine learning
--> parallel processing
--> neural networks
--> supervision / training
fMRI
-
disadvantages:
expensive, big, no movement (not alway possible in ppl with minimal vegetative state) etc..
fNIRS
--> temporal res good enough for motor cortex etc :3 !!
--> in-between of fMRI and MEG for temporal and spatial res
(spatial = >MEG but <fmri , and temporal = < MEG and >fMRI (then again firm can have sampling rate in milliseconds, but they don't do it cause movement artifacts :3 !!) (--> could also use equivalent of p300 ERP)
--> ratio = oxygenated / deoxy blood, the higher the ratio / the more oxygenated blood = the more activity :3 !!
-
-
-