Magnetic resonance imaging and the reduction of motion artifacts: review of the principles

Magnetic resonance (MR) imaging is a non-invasive diagnostic tool which is widely used nowadays. In this paper, the basic principles of MR imaging are explained and it is shown how images can be reconstructed in case of standard 2D Fourier Transform (2DFT) imaging. Several aspects of MR signal encoding are described. Unfortunately, motion of the patient during a magnetic resonance experiment often causes severe artifacts in the images. For example, in 2DFT imaging blurring and ghosting are seen and the appearance of motion artifacts remains one of the major drawbacks in MR imaging. Several methods to reduce motion artifacts in MR imaging have been proposed in the past. An overview of the principles on which these methods are based is given in this paper. Both post-processing methods and techniques that rely on gating or the use of alternative acquisition schemes such as projection reconstruction are discussed.     

Echo-planar imaging of the brain

In this review, the clinical utility of echoplanar techniques in MRI of the brain is discussed. Comparison of high-resolution EPI with SE/turbo-SE shows high image quality of EPI in the supratentorial brain. In the infratentorial region, however, susceptibility artifacts limit image quality. For the assessment of neuronal brain activation utilizing the intrinsic contrast of blood (BOLD), EPI has definite advantages over other techniques of functional MRI. Due to its superior temporal resolution and multislice capabilities, EPI allows for analysis of complex neuronal activation patterns. Diffusion imaging benefits from the lack of bulk motion artifacts and serves primarily to detect early stroke. Three methods of perfusion imaging (rel. blood volume, rel. blood flow) are discussed: the susceptibility artifact method (T2*), the relaxitivity method (T1), and the signal-labelling technique (STAR). Perfusion imaging may have a clinical impact in the assessment of brain tumors and cerebral ischemia.     

Two-dimensional multishot echo-planar coronary MR angiography

This work presents a two-dimensional (2D) multishot echo-planar imaging (EPI) technique for magnetic resonance angiography (MRA) of individual coronary arteries in a 17-heartbeat breath-hold. Conventional 2D and 3D segmented gradient-echo (GRE) coronary MRA requires repetitive excitation of the same slice or slab within each cardiac cycle, which can result in reduced blood signal and in motion artifacts. Two-dimensional multishot EPI can address these limitations by eliminating multiple excitations per cardiac cycle, using large flip-angle excitations, markedly reducing the data acquisition window, and performing oblique multislice 2D imaging. The goal of this study was to assess the feasibility of breath-hold 2D multishot EPI for multislice coronary MRA and to demonstrate its reliability by consistently acquiring high-quality images of the coronary arteries in a series of 16 volunteers.     

Elimination of artifacts in MRCP: technical consideration

Artifacts of MR cholangiopancreatography (MRCP) include fluid-filled gastrointestinal tracts, respiratory motion artifacts, spasm of Oddi's sphincter, vascular compression of bile ducts, overlapping of various anatomical structures and bright signal of surrounding fatty tissue. In this article, various technical considerations to eliminate these artifacts were described for the better imaging analysis of MRCP. The use of high-concentration ferric ammonium citrate (Ferriseltz) is recommended to reduce bright signal of fluid-filled gastrointestinal tracts. In case of long breathhold sequences, O2-inhalation study is useful to eliminate respiratory motion artifacts. Careful attention should be paid to the spasm of Oddi's sphincter and the vascular compression of bile ducts to avoid erroneous interpretation of MRCP findings.     

Analysis of dynamic brain imaging data

Modern imaging techniques for probing brain function, including functional magnetic resonance imaging, intrinsic and extrinsic contrast optical imaging, and magnetoencephalography, generate large data sets with complex content. In this paper we develop appropriate techniques for analysis and visualization of such imaging data to separate the signal from the noise and characterize the signal. The techniques developed fall into the general category of multivariate time series analysis, and in particular we extensively use the multitaper framework of spectral analysis. We develop specific protocols for the analysis of fMRI, optical imaging, and MEG data, and illustrate the techniques by applications to real data sets generated by these imaging modalities. In general, the analysis protocols involve two distinct stages: "noise" characterization and suppression, and "signal" characterization and visualization. An important general conclusion of our study is the utility of a frequency-based representation, with short, moving analysis windows to account for nonstationarity in the data. Of particular note are 1) the development of a decomposition technique (space-frequency singular value decomposition) that is shown to be a useful means of characterizing the image data, and 2) the development of an algorithm, based on multitaper methods, for the removal of approximately periodic physiological artifacts arising from cardiac and respiratory sources.     

Artifacts and pitfalls in MR imaging of the orbit: a clinical review

High-resolution magnetic resonance (MR) imaging of the orbit has become widely accepted as a valuable diagnostic technique. However, there are a number of artifacts and pitfalls associated with orbital MR imaging. Chemical shift artifacts may be induced by orbital fat or silicone oil used to treat retinal detachment. Motion artifacts are caused primarily by unavoidable globe motion during imaging. Artifacts due to a nonuniform magnetic field are particularly noticeable at air-tissue interfaces but may also be caused by incomplete fat saturation or highly magnetic materials near the orbit. Protocol errors may cause artifacts such as saturation, phase wraparound, truncation, shading, and partial-volume artifacts. This information can be used to improve orbital image quality and avoid misinterpretation of image artifacts. Use of fat saturation, silicone saturation, and careful patient screening for metal near the eyes and instruction to reduce motion can help reduce the occurrence of artifacts. In addition, optimal imaging technique is essential and should include use of proper surface coils, plane of section, and pulse sequences.     

In vivo echo-planar imaging of rat spinal cord

An integrated approach to echo-planar imaging of rat spinal cord in vivo with a small field of view (FOV) is presented. This protocol is based on a multishot interleaved echo-planar imaging (EPI) sequence and includes: 1) use of an inductively coupled implantable coil for improved signal-to-noise ratio (SNR); 2) three-dimensional (3D) automatic shimming of the magnetic field over the spinal cord; and 3) post-acquisition data processing using a multireference scan for minimizing image artifacts. Some of the practical issues in implementing this protocol are discussed. This imaging protocol will be useful in characterizing the spinal cord pathology using techniques that are otherwise time-consuming, such as diffusion tensor imaging.     

Assessment of myocardial perfusion hemodynamics using echo planar MR imaging

It was possible to obtain images for individual heart beats using single-shot Echo Planar Imaging(EPI), and changes of myocardial signal intensity could be assessed visually after GD-DTPA administration. Measurement of the same site in the myocardium on myocardial perfusion images for individual heart beats was facilitated by imaging during breath-holding, and accurate evaluation was possible. In patients with coronary artery disease, the site of myocardial infarction tended to show less increase in signal intensity than the normal myocardium, and could easily be distinguished from normal myocardium according to the change in signal intensity. In patients with atrial fibrillation, the signal intensity of the myocardium varied with each heart beat, and it was difficult to assess perfusion hemodynamics. Myocardial perfusion studies using EPI still present problems with respect to spatial resolution, but the myocardial perfusion hemodynamics for individual heart beats can be determined by preparing time/intensity curves. It is also possible to obtain information on cardiac morphology, wall motion, and myocardial metabolism in addition to perfusion data by combining myocardial perfusion studies with methods such as high speed cine MRI, tagging, or myocardial MRS. It is possible that this method will also be useful in studying myocardial viability.     

Brain functional MRI of the visual cortex with echo planar imaging

Brain functional MR imaging (fMRI) is a non invasive imaging method for detecting neural activity. We performed functional MRI of the visual cortex with gradient-echo echo planar imaging (GE-EPI) and spin-echo EPI (SE-EPI) using 1.5T MRI system. Visual stimuli was performed with a checkerboard patterns. Magnitude and temporal phase of correlation between each pixel's time-course and sine functions at the frequency of the stimulus was calculated. In all subjects, the activation area in visual cortex obtained from SE-EPI was smaller than that from GE-EPI. Temporal phase delay images from both GE-EPI and SE-EPI showed signal spread from the primary visual cortex to peripheral supplementary areas. Temporal phase analysis is important to discriminate the source of the hemodynamic response to neural activation in fMRI.     

Radial echo-planar imaging

A new ultrafast magnetic resonance imaging pulse sequence named radial echo-planar imaging (rEPI) is introduced. The sequence is based on a modification of the echo-planar imaging (EPI) sequence to scan k-space radially, in an attempt to combine the speed of EPI with the benefits of radial sampling. Like in EPI, all the desired lines in k-space are scanned consecutively in opposite directions. The unique feature of this new sequence, however, is that the orientation of the readout gradient is incrementally rotated, so that all the echoes are refocused through the center of k-space. Therefore, rEPI data are acquired in a polar grid, and image reconstruction can be done either by means of filtered back-projection or by regridding the data to a Cartesian matrix followed by 2D Fourier transform. First results show that rEPI images can be acquired with the same speed and signal-to-noise ratio of EPI images. rEPI images are also shown to be less sensitive to off-resonance effects than EPI images. Further studies are underway to investigate the usefulness of rEPI for spectroscopic imaging and applications affected by motion.     

Arachidonate-induced tyrosine phosphorylation of epidermal growth factor receptor and Shc-Grb2-Sos association

PURPOSE: Subdural grid arrays are used when seizure activity cannot be located by ictal scalp recordings and when functional cortical mapping is required before surgery. This study was performed to determine and compare the CT and MR imaging appearance of subdural EEG grids and to identify the types and frequency of associated complications. METHODS: We retrospectively reviewed the medical records and imaging studies of 51 consecutive patients who underwent 54 craniotomies for subdural EEG grid implantation with either stainless steel or platinum alloy contacts between June 1988 and September 1993. Twenty-two patients had both CT and MR examinations, 27 patients had CT only, and five patients had MR imaging only. All studies were assessed for image quality and degradation by the implanted EEG grids, for intra- and extraaxial collections, and for mass effect, with differences of opinion resolved by consensus. RESULTS: Subdural EEG grids caused extensive streak artifacts on all CT scans (corresponding directly to grid composition) and mild to moderate magnetic susceptibility artifacts on MR images. Sixteen associated complications were detected among the 54 patients imaged, including four significant extraaxial hematomas, four subfalcine or transtentorial herniations, two tension pneumocephali, two extraaxial CSF collections, two intraparenchymal hemorrhages, and one case each of cerebritis and brain abscess. In all but four cases, the detected complications were not clinically apparent and did not require specific treatment. There were no residual sequelae. CONCLUSION: Because of extensive streak artifacts, CT showed only gross complications, such as herniation and grid displacement by extraaxial collections. MR imaging artifacts were more localized, allowing superior evaluation of subdural EEG grid placement and associated complications.     

Endothelin receptors in rat pituitary gland

An algorithm is described for reducing ghost artifacts in echo planar imaging (EPI) using phase corrections derived from images reconstructed using only even or odd k-space lines. The N/2 ghost, that arises principally from time-reversal of alternate k-space lines, was significantly reduced by this algorithm without the need for a calibration scan. In images obtained in eight subjects undergoing EPI for auditory functional MRI (fMRI) experiments, N/2 ghost intensity was reduced from 10.3% +/- 2.1% (range: 7.9-14.1%) to 4.5% +/- 0.2% (range: 4.1-4.9%) of parent image intensity, corresponding to a percent reduction in ghost intensity of 54% +/- 9% (range: 43-65%), and the algorithm restored this intensity to the parent image. It provided a significant improvement in image appearance, and increased the correlation coefficients related to neural activation in functional MRI studies. The algorithm provided reduction of artifacts from all polynomial orders of spatial phase errors in both spatial directions. The algorithm did not eliminate N/2 ghost intensity contributed by field inhomogeneities, susceptibility, or chemical shift.     

Functional mapping of pain-related activation with echo-planar MRI: significance of the SII-insular region

Activation in numerous regions of the brain is likely to be involved in the complex neural network function of pain perception. To detect the cortical representation during nonpainful and painful stimuli, which were presented using electrical finger stimulation in six normal right-handed male volunteers, we performed echo-planar functional magnetic resonance imaging (fMRI). Using a 1.5-T MR system that scanned the supratentorial region of the brain, we obtained multislice BOLD-based functional MR images with single-shot gradient-echo echo-planar imaging (EPI). The data show that dispersed brain regions are activated during painful stimulation, and especially demonstrate the significance of the SII-insular region in pain perception.     

Studies on utility of MR T2-weighted images using multishot echo-planar imaging for hepatic mass lesions

MR T2-weighted images using multishot echo-planar imaging (EPI) and fast spin-echo (FSE) sequences were obtained in 22 patients with hepatic masses. Multishot EPI sequences included eight-shot breath-hold EPI and 16-shot EPI without breath-hold, while FSE sequences included nonfat-suppressed respiratory-triggered FSE, fat-suppressed respiratory-triggered FSE, and nonfat-suppressed breath-hold FSE. Signal-to-noise ratio, contrast-to-noise ratio and artifacts were compared between EPI and FSE images of 47 hepatic masses. In evaluating solid tumors, EPI provided image quality equal or superior to that of FSE, whereas in the evaluation of nonsolid tumors FSE showed better image quality than EPI. In conclusion, it was demonstrated that in the evaluation of hepatic solid tumors T2-weighted eight-shot breath-hold EPI can replace both nonfat-suppressed respiratory-triggered FSE and breath-hold FSE, and it was suggested that eight-shot breath-hold EPI can replace fat-suppressed respiratory-triggered FSE to reduce patient discomfort and increase examination throughput.     

New application of MRI by using echo planar imaging

EPI (Echo Planar Imaging) can provide human images with an acquisition time of less than 100 ms. EPI has received attention as a robust method for several applications, such as functional imaging, diffusion imaging, T2 imaging, and perfusion imaging. In addition to these, a couple of new EPI applications for brain functional imaging have recently been developed. One is spin targeting perfusion imaging, EPISTAR or FAIR, and the other is high-speed spectroscopic imaging, EPSI. Spin targeting perfusion imaging can detect inflow change alone during brain activation, contrast with bold contrast imaging that includes both O2 consumption and inflow effect. EPSI reduces signal acquisition time drastically. This review describes the recent advances in this field.     

On the masking of signals on immunoblots by cellular proteins

Functional magnetic resonance imaging (fMRI) is capable of detecting task-induced blood oxygenation changes using susceptibility sensitive pulse sequences such as gradient-recalled echo-planar imaging (EPI). The local signal increases seen in the time course are believed to be due to an increase in oxygen delivery that is incommensurate with oxygen demands. To help isolate the sources of functional signal changes, the authors have incorporated various forms of diffusion weighting into EPI pulse sequences to characterize the apparent mobility of the functionally modulated protons. Results suggest that the majority of the functional signal at 1.5 T arises from protons that have apparent diffusion coefficients that are approximately four or five times higher than that of brain tissue. This implies that significant functional signal sources are either protons within the vascular space or protons from the perivascular space that is occupied by cerebrospinal fluid.     

Functional burst imaging

A quiet magnetic resonance (MR) imaging technique for detecting changes in cerebral activity functions is presented. This single-shot method, functional Burst imaging (FBI), combines elements of Burst imaging with an offset technique known as asymmetric spin echo (ASE). The FBI sequence has the unique feature of being nearly silent, because of the low number of gradient switching steps involved. Furthermore, this approach has the key advantage that it can be implemented on conventional MR systems. Established auditory and visual paradigms were used to evaluate whether FBI can detect changes in cerebral activity using a 1.5 Tesla MR system. In a second set of experiments, the FBI technique was used to evaluate cerebral activity changes during different sleep stages in humans. The results obtained demonstrate that the FBI sequence provides an alternative approach for functional imaging of brain activity in primary and secondary sensory areas of the human brain. Furthermore, in using this quiet MR technique, it was possible to scan continuously during different stages of human sleep without acoustic noise perturbation.     

Multishot echoplanar MR imaging of the female pelvis: comparison with fast spin-echo MR imaging in an initial clinical trial

OBJECTIVE: The purpose of this study was to apply multishot echoplanar MR imaging (EPI) to the female pelvis and compare image quality with that of fast spin-echo MR imaging. SUBJECTS AND METHODS: Twenty-one patients with suspected pelvic disease and five healthy female volunteers were prospectively examined. MR imaging was obtained using an EPI-capable 1.0-T imager and a pelvic phased-array coil. Axial EPI and fast Spin-echo sequences were obtained at identical image locations in each patient and volunteer. Spin-echo EPI images were obtained using a multishot number of 16. Acquisition time for each EPI sequence was 2 min 10 sec. Fat-suppressed T2-weighted fast spin-echo images were obtained in 2 min 12 sec. Paired EPI and fast spin-echo MR images were independently evaluated by three reviewers. RESULTS: Delineation of the perivaginal and parametrial venous plexus on EPI was rated superior to the fast spin-echo images in 62 (77%) of 81 cases. On EPI, uterine zone anatomy and ovary visualization were judged to be inferior in 44 (56%) of 78 cases and in 18 (33%) of 54 cases, respectively. For delineation of uterine abnormalities, including leiomyoma and adenomyosis, both sequences performed almost equally well. However, ovarian cystic lesions were revealed more precisely by the fast spin-echo sequence. An overall fat-suppression effect was seen on EPI images in 62 (77%) of 81 cases. CONCLUSION: Multishot EPI cannot replace fast spin-echo sequences for imaging the female pelvis; however, because EPI has a potent fat-suppression effect and heavily T2-weighted contrast, EPI sequences can be a valuable adjunct to routine examination.     

MR cholangiopancreatography with single-shot fast-spin-echo sequences]

Imaging quality of MR cholangiopancreatography (MRCP) has recently made a great advance, and MRCP plays an important part in the diagnosis of pancreticobiliary diseases. To obtaining excellent-quality MRCP images, three conditions are required: fluid-to-background contrast, high spatial resolution, and suppression of respiratory motion artifacts. Respiratory motion artifacts, whose suppression is especially important for MRCP, can be controlled by the use of breath-holding, signal averaging, or respiratory triggering. These images are implemented in either single-slice technique or multi-slice technique. We have performed MRCP prior to endoscopic retrograde cholangiopancreatography (ERCP) in more than 100 patients. MRCP images were obtained as maximum-intensity-projection (MIP) reconstruction images and thick-slice projection images by using single-shot fast-spin-echo sequence, and MIP reconstruction images by using respiratory-triggered fast-spin-echo sequence. We reviewed MRCP imaging with single-shot fast-spin-echo sequences.     

MR angiography of the portal venous system: techniques, interpretation, and clinical applications

Magnetic resonance (MR) angiography is a noninvasive means of assessing the portal venous system that has potential advantages over currently used modalities. Time-of-flight and phase-contrast MR angiography are useful techniques that differ fundamentally in their means of data acquisition but are comparable in their ability to demonstrate normal anatomy as well as abnormalities of the portal venous system. Occasionally, artifacts caused by respiratory motion, implanted metallic devices or surgical clips, in-plane saturation, or areas of complex flow are seen at MR angiography of the portal venous system. However, most artifacts can easily be identified as such and either remedied or ignored. In addition, the suppression of signal from surrounding soft tissues may result in poor detection of parenchymal lesions. The utility of standard projection angiograms and source images can be increased through the use of intravenously administered contrast material and postprocessing techniques such as partial-volume maximum intensity projection reconstructions and shaded surface renderings. In addition to providing information on portal venous anatomy and portosystemic collateral vessels, MR angiography of the portal vein has clinical application in portal venous thrombosis and stenosis, liver transplantation, and the evaluation and planning of surgical and transjugular intrahepatic portosystemic shunts.