2009 Max Halvorson

Back to Psych 204 Projects 2009

Differences Between Spin-Echo and Gradient-Echo Imaging

Spin-Echo and Gradient-Echo imaging are two popular methods of obtaining fMRI data that vary along various dimensions: the pulse sequence used to generate and obtain the signal, signal-to-noise ratio, and sensitivity to large blood vessels, to name a few. This study examines the differences between the two by looking at a data set from the VISTA lab. Overall differences in signal-to-noise, areas of high and low signal, and distortions are examined. Furthermore, hV4 and other areas of visual cortex are considered and compared between the two types of scans.

• Note: another goal of this study was to familiarize the author (who has no experience viewing fMRI data) with fMRI data, analysis techniques, software, and different types of data.

Background

Gradient-Echo Imaging

Gradient-Echo images are generated by an applied gradient-followed by an RF pulse sequence to excite slices one by one and collect data from each. Gradient-Echo images have a stronger overall signal than Spin-Echo images and the overall signal-to-noise ratio is higher. However, Gradient-Echo images are prone to large distortions from large blood vessels, sinuses, and other inhomogeneities.

Spin-Echo Imaging

Spin-Echo images are generated by a 180-degree pulse following the applied gradient. This pulse realigns the dephasing spins and gives another shot at capturing data. Spin-Echo images are weaker in signal and signal-to-noise ratio than Gradient-Echo images. However, Spin-Echo images should be less sensitive to distortions from large blood vessels and sinuses.

Vessel Distortions

Spin-Echo imaging can be used to refocus the loss of phase coherence and eliminate the large-vessel signal. For smaller vessels, the gradient changes rapidly over space relative…Loss of phase coherence cannot be recovered by Spin-Echo imaging (from Huettel et al. text).

Purpose

Although Gradient-Echo images have a stronger overall signal and signal-to-noise ratio and generally show more in most brain regions, Spin-Echo images are theorized to avoid distortions due to large blood vessels and inhomogeneities. The hope of this study was to find areas in which Spin-Echo images could reveal things in visual cortex that Gradient-Echo could not. For example, areas like hV4...

Methods

The data were obtained from Jon Winawer through VISTA lab.

MR Analysis

The MR data were analyzed using mrVista software tools. Features used included correlation analyses, traveling wave analyses, mean maps, phase-projected coherence maps, and time series plots.

Pre-processing

All data were slice-time corrected, motion corrected, and repeated scans were averaged together to create a single average scan for each subject. Pre-processing was done by Jon.

Results

(Figure 1) Slice Origin

(Figure 4) Raw Time Series of Areas V1-V3 (bilateral)

(Figure 6) Average Time Series of Right hV4

(Figure 5) Average Time Series of Areas V1-V3 (bilateral)

(Figure 7) Single Cycle of Right hV4

(Figure 2) Gradient-Echo vs. Spin-Echo

(Figure 3) Major Distortions in Gradient-Echo vs. Spin-Echo

(Figure 8) Sinus ROI and nearby visual cortex

(Figure 9) Gradient-Echo FFT

(Figure 10) Spin-Echo FFT

Conclusions

Overall

At first glance, one notices the difference in overall signal and signal-to-noise between the two scans as indicated by the brighter colors in the Gradient-Echo scan and the greater contrast between areas of signal areas of no signal (Figure 2). Another striking difference is seen in the type of distortions caused by the ear canals in the Gradient-Echo (absence of signal in that area) and the Spin-Echo ("bubbles" of signal built up over that area) scans (Figure 3). Furthermore, the phase, as shown by the FFT of each scan, of the two signals is right in-sync, with both peaking at the same time (Figures 9 & 10).

The next thing to note is the overall similarity in signal and phase between the Spin-Echo and Gradient-Echo signals for V1-V3 (Figures 4 & 5). One also notices that the Spin-Echo scan has a much lower overall signal throughout. Though the Spin-Echo scans have a weaker signal, they seem to reflect similar data for most areas of visual cortex (analyses of cortex other than V1-V3 not shown). In areas like the ear canals where the distortions produced differ in character, Spin-Echo imaging doesn't do a better job either.

hV4

In the areas of interest, specifically left and right hV4 near the sinus, Spin-Echo imaging did no better than Gradient-Echo imaging in gathering a coherent signal from hV4. Furthermore, surrounding areas have distortions in Spin-Echo scans which are absent in Gradient-Echo scans (Figure 8). It seems that Spin-Echo scans, at least in this case, provide no useful additional information about hV4.

Implications

Further research might examine other methods of obtaining functional data in this region (e.g. different pulse sequences). Also, other regions of the brain remain to be examined. Perhaps there are areas of (visual and non-visual) cortex in which Spin-Echo scans offer a distinct advantage.

References

Software: mrVista

Winawer J, Horiguchi H*, Sayres R*, Amano K, Wandell BA. (In Press) Mapping hV4 and ventral occipital cortex: The venous eclipse. Journal of Vision.