0.25
0.5
0.75
1.25
1.5
1.75
2
Pattern-Information Analysis: Columnar Sensitivity, Stimulus Decoding, and Computational-Model Testing
Published on Dec 03, 20122849 Views
Pattern-information fMRI has become a popular method in neuroscience. The technique is motivated by the idea that spatial patterns of fMRI activity reflect the neuronal population codes of perception,
Related categories
Chapter list
Pattern-information fMRI00:00
Overview00:18
Part A: Learning objectives02:17
Why fMRI?03:45
Spatial and temporal resolution03:58
Number of channels and coverage06:02
Similar information rates07:28
Why pattern-information fMRI?08:31
Activation analysis (1)08:47
Activation analysis (2)09:26
Pattern-information analysis10:51
Filling the explanatory gap16:03
theory - experiment (1)17:38
theory - experiment (2)20:57
Why decode fMRI?21:55
“Decoding” stimuli from brain responses21:58
Linear pattern-classifier analysis (1)24:06
Linear pattern-classifier analysis (2)25:19
Nonlinear classifiers?26:37
Decoding or encoding?31:50
from Kay & Gallant (2009)33:56
“Predicting” brain responses from stimuli34:48
Potential title claims35:54
Model direction has no neuroscientific implications (1)37:08
Model direction has no neuroscientific implications (2)37:10
Why test encoding models with fMRI?38:11
decoding - encoding (1)39:00
decoding - encoding (2)43:02
Predicting brain responses for novel stimuli45:30
“Predicting” novel stimuli from brain responses (1)49:25
“Predicting” novel stimuli from brain responses (2)52:51
Which distinctions between stimuli are reflected and which are lost in the population code?53:51
Inferior temporal (IT) representational space (1)55:24
Inferior temporal (IT) representational space (2)56:15
Inferior temporal (IT) representational space (3)57:43
Comparing brain and model representations (1)58:35
Comparing brain and model representations (2)01:01:19
Model dissimilarity matrices (1)01:03:19
Model dissimilarity matrices (2)01:03:20
Human IT (1)01:04:49
Human IT (2)01:07:23
Human IT (3)01:08:12
theory - experiment01:09:17
What neuroscientific implications do these different types of pattern-information result have?01:09:35
Kriegeskorte (2011)01:09:51
Why should your analysis be at once data- and hypothesis-driven? (1)01:14:15
Why should your analysis be at once data- and hypothesis-driven? (1)01:14:19
Hypothesis-driven approach: distinguish two models01:19:07
Data- and hypothesis-driven approach!01:20:47
Quiz01:23:16