FMRI Artifacts

Magnetic resonance imaging has recently found application in mapping human brain function. The contrast mechanism used is the dependence of the signal intensity on the local tissue concentration of deoxyhemoglobin. The relationship is approximately logc{S/S0} oc — [deoxyHb]. During neuronal activity, the brain locally consumes more oxygen, but because of an even greater increase in local blood flow the venous blood becomes less deoxygenated and the signal increases. These subtle changes in signal (1-5%) can be detected using gradient-echo pulse sequences [23,36]. Because the signal changes are small, it is desirable to collect many MRI volumes and to alternate several times between the neurological task stimulus and the control stimulus. For this reason fast imaging sequences such as EPI and spiral imaging are used, capable of sampling up to 10 slices of the brain per second.

6.4 Signal Aliasing

Another example of artifact that is frequently encountered in MRI data sets is the effect of signal aliased from regions outside the field-of-view back into the image. Aliasing is rare in the frequency-encoding (readout) direction of the image, since analog or digital filters are often used to filter frequencies outside the desired field of view. Conversely, aliasing in the phase-encode direction is quite common.

The Nyquist sampling theory dictates that a signal that is sampled with a dwell time DW can only accurately represent signals up to a maximum frequency + ^N, where fN = 1/(2 xDW). In the absence of filtering, frequencies outside this Nyquist threshold will fold back into the spectrum such that a signal of frequency fN + will be represented in the spectrum at position — fN + 8f. This often happens in the phase-encode direction of the MRI image when signal extends outside the phase-encode direction field of view. Three examples of this are shown in Figs 8e, 8f, and 8g. Figure 8e shows the effect of positioning the sample off-center with respect to the MRI field of view. Signal from the bottom of the image has aliased to the top of the field of view. Clearly, signal can alias to a sufficient extent that it overlaps the main image.

7.1 Echo Planar Imaging and Spiral Imaging Artifacts

EPI is the most common pulse sequence used in functional MRI studies. Developed in 1977, it is only in recent years that improvements in scanner hardware have enabled EPI to be used widely. Impressive speed is attainable with EPI. Up to 10 slices per second maybe sampled, although the pixel resolution is usually limited to 64 x 64 pixels or 128 x 128 pixels. However, EPI suffers from two significant artifacts, as described next.

Nyquist Ghost

The principle of scanning fc-space in a single shot was described in Section 3.2 and in Fig. 3b. In EPI all the lines of fc-space are sampled following a single excitation of the spin system. As is evident in Fig. 3b half the fc-space lines are acquired under a positive readout field gradient (left to right) and half are collected under a negative readout field gradient (right to left). This is in contrast to conventional imaging, where all lines are collected under a field gradient of the same polarity (Fig. 3a). During image reconstruction the lines collected right to left are time reversed to match the lines collected left to right. Despite this, discrepancies in phase, timing, and amplitude will remain between the two sets of fc-space lines. Since, for snapshot EPI, every alternate line in fc-space is collected with a gradient of opposite polarity, any differences in timing or receiver response between the odd and even lines will alternate with the Nyquist frequency. This induces a ghost of the main image that is displaced by half the field-of-view in the phase-encode direction.

Figure 9a shows an example of an EPI ghost in a simulated phantom. An incorrectly set receiver delay results in systematic phase and timing differences between the odd and even fc-space lines. Figure 9b shows the result of calibrating and correcting for the phase difference between the odd and even fc-space lines. This is often carried out by collecting a reference scan without any phase encode gradient pulses [6,20,45]. It is also possible to apply a correction without a reference scan [7]. Use of a reference scan provides a substantial correction, but it is difficult to achieve ghosts that are less that 5% of the main image intensity. Finally, if the complex time domain data is available from the scanner, it is possible to perform an empirical post-hoc minimization of the ghosting by searching for appropriate zeroth; and first-order phase shifts in the semiFourier transformed data S(x, fcy) that minimize the Nyquist ghost in p(x, y).

Field Inhomogeneity Effects

Field inhomogeneity effects in EPI data were discussed in Section 3.2. The effect of field inhomogeneity is to distort both

FIGURE 9 Illustration of the combined effects of Nyquist ghost and geometric distortion in an EPI pulse sequence. The computer-generated phantom is a disk with a regular grid running through it. A poor field homogeneity with a (x2 — y2 + xy) dependency was also simulated. (a) The result of simply time reversing alternate lines in fc-space and Fourier transforming. Significant geometric distortion (from the poor shim) and Nyquist ghosting (from misset timing) are evident. (b) The result after phase correcting the semi-Fourier transformed data with a reference scan (collected without phase-encode gradients). The Nyquist ghost is substantially suppressed, although the geometric distortion remains. The details of the phase correction strategy used may partially correct for geometric distortions, but will not completely remove them.

the geometry of the sample and also the intensity. The most prominent effect is usually the geometric distortion, which will cause a static magnetic field inhomogeneity, ABz (x, y), to mislocate pixels in the phase-encode dimension of the image according to y = y0 + yABz(x, y)tpc, where is the effective phase-encode dwell time (approx. NDW). This mislocation can be up to 10% of the field of view. The most important implication of this mislocation is that it must be allowed for when registering EPI data with images collected using a more conventional pulse sequence. Either a nonlinear spatial warp must be allowed for in registration, or information from the measured field inhomogeneity distribution must be used [22].

Spiral imaging sequences are affected differently by field inhomogeneity. For spiral pulse sequences, fc-space is sampled in a spiral that starts from the center of fc-space and spirals out until reaching the desired fcmax radius. A regridding procedure is next performed to interpolate the acquired points onto a rectilinearly sampled equidistant grid. A 2D FFT then provides the image. The effect of local field inhomogeneities is to broaden the point-spread function of the signal in both Fourier dimensions. The spiral image therefore becomes locally blurred. Strategies to correct for this problem include using knowledge of the measured field inhomogeneity distribution to provide a correction [33] or deducing the distribution from the data to a low polynomial order [21,28]. Some local spatial blurring will inevitably remain, as well as some streaking artifacts introduced during fc-space regridding.

7.2 Physiological Noise Effects

The magnitudes of signal changes observed in the various functional MRI techniques are quite small. Even for functional tasks involving primary sensory or primary motor areas, the fractional change in MRI signal attributable to neuronal activity is usually less than ~ 4%. For more subtle cognitive tasks, for example, working memory tasks, or for single-event fMRI studies, the magnitude ofthe signal changes can be on the order of 1% or less. Once the scanner hardware is optimized and head motion of the subject has been corrected [15,46], the most significant residual source of noise (or more precisely the temporal instability) in the images is the effect of physiological processes. These mainly consist of respiratory and cardiac effects.

Often, the largest source of physiological artifact in fMRI time series is caused by respiration. This is revealed as signal changes that correlate with the breathing cycle. Experiments that have been performed to determine the origin of respiration-related artifacts have indicated that the effect is largely ascribable to small magnetic field shifts induced in the brain, which in turn are caused by gross magnetic susceptibility changes as the lungs in the chest cavity expand and contract. These magnetic field shifts have been measured [34, 35, 47] and have values for maximum inspiration minus maximum expiration of 0.01 ppm at the most superior part of the brain, increasing to 0.03 ppm at the base of the brain (closest to the chest cavity).

The cardiac-related noise effects are largely caused by motions of the brain that in turn result from cerebral blood volume and pressure fluctuations [14]. Poncelet et al. [40] and Enzmann and Pelc [13] used motion-sensitive MRI methods with high temporal resolution to characterize the bulk motions of the brain that correlate with the cardiac cycle. The motions detected using these techniques showed that brain structures close to the brain stem moved with excursions of up to 0.5 mm over the cardiac cycle, whereas cortical regions of the brain moved with excursions that are less than ~ 0.05 mm. This implies that bulk motion of the brain caused by cardiac contraction could be a source of artifact, particularly in the deeper brain structures. Also, the varying effects of inflow of fresh blood spins into the slice of interest over the heart cycle can lead to signal fluctuations (see Section 4.1).

Numerous correction strategies have been proposed to minimize physiological signal fluctuations in fMRI data sets. Methods based on navigator echo approaches have been used [18,24,34,35]. Another approach is to externally monitor the physiological processes themselves and to fit a low order polynomial function to the "unit cycle" that describes the point-by-point phase and amplitude modulation of the signal throughout the cardiac and respiratory cycles. Those effects can then be removed by vector subtraction [19]. A substantial improvement in the statistical power of the data can be realized using these methods.

Was this article helpful?

0 0

Post a comment