In this study 1 anatomical and 8 functional scans assessed , 4 of them are motor task(finger tapping), the rest 4 of them are visual task flickering checker board for left and right visual hemi-field.
functional scans are acquired with following parameters.
f1) Visual Task 3x3x3 TR2 TE35 101TR with EP2D
f2) Motor Task 3x3x3 TR2 TE35 101TR with EP2D
f3) Visual Task 3x3x3 TR1 TE35 202TR with CMRR EP2D MB4
f4) Motor Task 3x3x3 TR1 TE35 202TR with CMRR EP2D MB4
f5) Visual Task 3x3x3 TR2 TE35 101TR with CMRR EP2D MB4
f6) Motor Task 3x3x3 TR2 TE35 101TR with CMRR EP2D MB4
f7) Visual Task 3x3x3 TR0.5 TE35 404TR with CMRR EP2D MB7
f8) Motor Task 3x3x3 TR0.5 TE35 404TR with CMRR EP2D MB7
==============================
GLM Results for Visual Task
==============================
The figure above corresponds for the following functional tasks respectively.
Functional 1) Visual Task //3x3x3 voxels // TR 2seconds // 101 volume with conventional EPI
Functional 5) Visual Task //3x3x3 voxels// TR 2seconds // 101 volume with Multiplexed EPI with MB 4
Functional 3) Visual Task //3x3x3 voxels// TR 1second // 202 volume with Multiplexed EPI with MB 4
Functional 7) Visual Task //3x3x3 voxels// TR 0.5second // 404 volume with Multiplexed EPI with MB 7
==============================
GLM Results for Motor Task
==============================
The figure above corresponds for the following functional tasks respectively.
Functional 2) Motor Task //3x3x3 voxels // TR 2seconds // 101 volume with conventional EPI
Functional 6) Motor Task //3x3x3 voxels// TR 2seconds // 101 volume with Multiplexed EPI with MB 4
Functional 4) Motor Task //3x3x3 voxels// TR 1second // 202 volume with Multiplexed EPI with MB 4
Functional 8) Motor Task //3x3x3 voxels// TR 0.5second // 404 volume with Multiplexed EPI with MB 7
==============================
cbICA Results for Visual Task
==============================
The figure above corresponds for the following functional tasks respectively.
Functional 1) Visual Task //3x3x3 voxels // TR 2seconds // 101 volume with conventional EPI
Functional 5) Visual Task //3x3x3 voxels// TR 2seconds // 101 volume with Multiplexed EPI with MB 4
Functional 3) Visual Task //3x3x3 voxels// TR 1second // 202 volume with Multiplexed EPI with MB 4
==============================
cbICA Results for Motor Task
==============================
The figure above corresponds for the following functional tasks respectively.
Functional 2) Motor Task //3x3x3 voxels // TR 2seconds // 101 volume with conventional EPI
Functional 6) Motor Task //3x3x3 voxels// TR 2seconds // 101 volume with Multiplexed EPI with MB 4
Functional 4) Motor Task //3x3x3 voxels// TR 1second // 202 volume with Multiplexed EPI with MB 4
Comments & Conclusion
* M-EPI with multiband factor 4 and TR 1second performed best in this comparison by defining the active region more specifically , this is because ; sequence collects twice more data in the same acquisition time which doubles the sampling of the time courses thus uses the dynamic range of correlation analysis better and resulted as an increased spatial specificity of ROI.
M-EPI with further increased multiband factors were not resulted well because of to much geometric distortions in functional images.
*It could be said that ICA resulted better in visual task, that is because visual task is hard to follow and long eye fixations are impractical .Check the 3rd functional scan; there is depression of BOLD signal in 2nd cycle (around 40th volume).This bold depression occurred because of subject could not conduct the task as desired at that time. This problem resulted as a decreased correlation in GLM analysis and worser spatial maps in ROI.
However cbICA uses that time course as a component and resulted as better defined ROI.
*Cortex based mask that used in cbICA analysis was prepared by the help of FreeSurfer segmentation. vbICA analysis results are not good to compare with GLM. Cortex based ICA (cbICA) results are far better than vbICA .Because of decreased amount of noise sources in selected time courses and applied GM mask on cortex.
