Compressed Sensing for fMRI: Feasibility Study on the Acceleration of Non-EPI fMRI at 9.4T

Figure 12

Example showing the effect of on reconstruction using Algorithm 1 (k-t FOCUSS with temporal FT). Residual error with variation using Algorithm 1 on (a) bSSFP PC0, (b) bSSFP PC1, (c) bSSFP PC2, (d) bSSFP PC3, and (e) GRE data is shown for each different . Plots are obtained from reconstruction of downsampled data using during testing phase of a representative rat, and similar results were obtained from different sampling patterns and different subject rats. Notice that residual error reaches convergence as increases. The chosen value of 30 is represented in each plot as a red dashed line.

(a)

(b)

(c)

(d)

(e)