Dear Dr YAN
i hope you do will.
1-Video is very fantastic ,but really i want reference for thioretical information of clusters.
2- if i have this cluster:
Cluster 1
Number of voxels: 60
Peak MNI coordinate: -36 -78 -45
Peak MNI coordinate region: // Left Cerebellum // Cerebellum Posterior Lobe // Inferior Semi-Lunar Lobule // undefined // undefined // Cerebelum_Crus2_L (aal)
Peak intensity: 4.2667
# voxels structure
60 --TOTAL # VOXELS--
57 Left Cerebellum
57 Cerebellum Posterior Lobe
42 Cerebelum_Crus2_L (aal)
24 Inferior Semi-Lunar Lobule
18 Pyramis
11 Tuber
3 Uvula
2 Cerebelum_Crus1_L (aal)
1 Cerebellar Tonsil
my questions :
2.1-the nomber of these voxels (42Cerebelum_Crus2_L+2 Cerebelum_Crus1_L) are the activated voxels and (60-44) not activated?
2.2-if the number of voxel is very small like 2 Cerebelum_Crus1_L (aal), is this region active region or not?
2.3- if the region repeated in two or more clusters what does mean that?
2.4- if I set cluster size with 13 voxels, can i take the 2 Cerebelum_Crus1_L (aal) as activation region?
2.5-Based on what these regions included in this cluster?
2.6- Peak intensity: 4.2667: is this the peak value for the region that have maximum voxels in the cluster?
3- I have P-value after FDR correctin =0.0037 , rmm=6 and 4mm gaussian smmoth , how many voxels in cluster size?
Are P- values in this image after FDR correction or before?
4- if I have this cluster:
Cluster 4
Number of voxels: 257
Peak MNI coordinate: -48 6 -33
Peak MNI coordinate region: // Left Cerebrum // Temporal Lobe // Middle Temporal Gyrus // White Matter // undefined // Temporal_Mid_L (aal)
Peak intensity: 4.9106
# voxels structure
257 --TOTAL # VOXELS--
88 Temporal_Inf_L (aal)
60 Fusiform_L (aal)
20 Temporal_Pole_Mid_L (aal)
14 ParaHippocampal_L (aal)
5 Temporal_Mid_L (aal)
4.1-if i set cluster size to 13 voxels which regions choose as active and which of them noise regions?
4.2 -Can i take only the peak region from cluster or can i take more than one region in adition to the peak region?
thanks for your help.
Forums
2.1. Those are just labels,
2.1. Those are just labels, that means some number of voxels are labeled to some area name. They are NOT exclusive. All 60 voxels are significant.
2.2. The 2 voxels are significant at the cluster level. However, that doesn't mean you have enough confidence to interpret it as Cerebelum_Crus1_L is significant.
2.3. The are not exclusive, one voxel can be labeled to multiple are names.
2.4. See 2.2.
2.5. xjview.
2.6. Yes.
3. I didn't get the point. FDR only gives you a threshold of p value.
4.1. I guess you can interpret it as inferior temporal cortex and fusiform cortex.
4.2. As many as you think reasonable.