Table 1.
Imaging parameters for task-based 2-D EPI, 3-D PRESTO and 3-D FFE fMRI sequences.
Table 2.
Imaging parameters for 2-D EPI, 3-D PRESTO and 3-D FFE fMRI sequences with high temporal resolution and navigator pulse.
Table 3.
Imaging parameters for 2-D EPI, 3-D PRESTO and 3-D FFE fMRI sequences for comparing signal and noise in the midbrain and cortex.
Figure 1.
TSNR in EPI, FFE and PRESTO sequences comparing different analysis methods in task-based scans.
Panel A 2-D EPI, panel B, 3-D FFE and panel C, 3-D PRESTO. Analysis methods include: R0 = no RETROICOR correction, R1 = only RETROICOR, PR0_R1 = no phase regression but with RETROICOR, and PR1_R1 = phase regression and RETROICOR. Regions of interest in the midbrain include the Substantia Nigra (SN), and the Ventral Tegmental Area (VTA). TSNR values in both left and right sides of the ROI are averaged; N = 5 in each sequence. TSNR values are under 10 for the SN and under 20 for the VTA irrespective of the analysis technique used.
Figure 2.
Plots of brainstem motion during cardiac systole using a retrospective cardiac gated scan.
Plots display displacement and average velocity of a region of interest from the original position. These measurements are calculated along the cardiac cycle. LR = Left-Right direction; AP = Anterior-Posterior direction; HF = Head-Foot direction. Regions of interest include SN = Substantia nigra; VTA = Ventral Tegmental Area; L/R_SN = left/right Substantia Nigra; L/R_VTA = left/right Ventral Tegmental Area, and the Midbrain. Greatest displacement for the midbrain observed in the anterior and inferior directions.
Figure 3.
Spatial power spectrum maps of noise in midbrain areas within respiratory and cardiac frequency ranges.
Power spectrum maps in the midbrain for high temporal resolution PRESTO (Panel A, B) and EPI (Panel C, D) scans sequences. Only representative slices covering the midbrain are displayed. Color scale indicates percentage of noise in respiratory and cardiac frequency ranges with respect to total noise in each voxel. Respiration frequency range = 0.08–0.25 Hz, and cardiac frequency range = 0.8–1.1 Hz. P denotes the posterior part of the brain. Physiological fluctuations in the midbrain (cardiac or respiratory) comprised less than half of the total noise variance in resting state data.
Figure 4.
Spatial TSNR maps using a 1-D phase navigator pulse in 3-D multi-shot sequences.
Only representative slices covering the midbrain are displayed. Color bar indicates TSNR values. P denotes the posterior part of the brain. Maps indicate a slight decrease in midbrain TSNR after 1-D navigator correction for both PRESTO and FFE data.
Figure 5.
TSNR values from analysis with regressor of no interest.
TSNR values in analysis with regressor of no interest using the white matter area of the cerebral peduncles (adjacent to the SN) and anterior cerebellum (posterior to the superior colliculi). The different analyses compared are: R0 = no RETROICOR, R1 = RETROICOR only, WM_R0 = WM without RETROICOR, WM_R1 = WM with RETROICOR, ACb_R0 = ACb without RETROICOR, ACb_R1 = ACb with RETROICOR. Average TSNR over five individuals in FFE data are displayed. Regions of interest in the midbrain include the Substantia Nigra (SN) and the Ventral Tegmental Area (VTA). TSNR values in both left and right sides of the ROI were averaged and N = 5. Values are under 10 for the SN and under 15 for the VTA irrespective of the analysis technique used.
Figure 6.
Partial correlations between midbrain ROIs in band-pass filtered fMRI data.
Partial correlations between midbrain ROIs are displayed accounting for shared temporal variance with neighboring white matter in resting state filtered data (N = 6). 2-D EPI scan is shown in panel A, 3-D FFE in panel B and 3-D PRESTO is in panel C. Dots represent the spread of correlation values across participants. Blue dots denote the homologous left and right SN, red dots denote the homologous left and right VTA, and black dots denote the non-homologous ROI pairs. The green dashed line denotes the mean correlation across runs and across participants for each ROI pair. Statistical analysis demonstrated evidence for differences between ROI pairs for FFE (repeated measures ANOVA, F(5,25) = 8.905, p = 6×10−5), and PRESTO (F(5,25) = 5.601, p = 0.001), but not EPI (F(5,25) = 0.684, p = 0.640). In post-hoc tests, the differences were primarily between homologous vs. non-homologous pairings (shown in Table 4). Greater partial correlations were observed between homologous SN and VTA compared to the non-homologous pairs, particularly for the 3-D sequences.
Table 4.
Pairwise differences between homologous midbrain ROI pairs (left and right SN, left and right VTA) and their non-homologous counterparts in 2-D EPI, 3-D FFE and 3-D PRESTO data.
Figure 7.
ROI seed based functional connectivity maps in the midbrain for 3-D sequences.
Functional connectivity maps for FFE (Panel A) and PRESTO (Panel B) scans in a representative participant. An uncorrected voxel level threshold of p<10−6 (cluster size = 30 voxels) was used for the FFE scans and an uncorrected voxel level threshold of p<10−10 (cluster size = 30 voxels) was used for the PRESTO scans. Seed regions of interest in the midbrain include left Substantia Nigra (L_SN), right Substantia Nigra (R_SN), left Ventral Tegmental Area (L_VTA), and right Ventral Tegmental Area (R_VTA). Color bar represents T-statistic. FFE voxel size = 1.33 mm3 and PRESTO voxel size = 2.5 mm3. Axial midbrain slice sections are displayed at the level of the superior colliculus. N = 6 in each sequence type. L denotes the left side of the brain and P denotes the posterior portion of the brain. Data were motion corrected, put through RETROICOR, and band-pass filtered between 0.01 and 0.1 Hz. Robust bilateral spatial distribution of functional connectivity was observed in most ROIs.