MR Image Quality Research Articles

Dynamic Contrast-Enhanced Magnetic Resonance Images of the Kidney

In this article, an improved method to segment the renal cortex and medulla and to eliminate the influence of kidney motion produced by respiration is proposed. On the basis of an adaptive threshold estimated from the mean gray levels of the grown regions, a region-growing algorithm is presented to produce a three-dimensional (3-D) kidney contour and to segment the renal structures. Moreover, a shell mask of kidney margin is proposed to realize a coarse matching so as to eliminate the image translation, which makes the processing simple and direct in spatial processing without any image transform computations, and a correlation computation can be implemented with great efficiency. Then, a refined matching with a mask of cortex is completed through a 3-D correlation algorithm to ensure the accurate registration of the images in different phases. Comparing with the global mask including the whole kidney, both the shell mask and the cortex mask significantly contribute to decreasing the matching errors for images with nonuniform intensity signals, which much improves the registration quality of renal MR images. In this way, the effect of the respiration motions can be eliminated so that the intensity measurement in different phases becomes accurate within the respective structures.

Image‐based coronary tracking and beat‐to‐beat motion compensation: Feasibility for improving coronary MR angiography

A method to reduce the effect of motion variability in MRI of the coronary arteries is proposed. It involves acquiring real-time low-resolution images in specific orthogonal orientations, extracting coronary motion from these images, and then using this motion information to guide high-resolution MR image acquisition on a beat-to-beat basis. The present study establishes the feasibility and efficacy of the proposed approach using human motion data in an offline implementation, prior to future online implementation on an MRI scanner. To track the coronary arteries in low-resolution real-time MR images in an accurate manner, a tracking approach is presented and validated. The tracking algorithm was run on real-time images acquired at 15-20 frames per second in four-chamber, short-axis, and coronal views in five volunteers. The systolic and diastolic periods in the cardiac cycles, computed from the extracted motion information, had significant variability during the short time periods typical of cardiac MRI. It is also demonstrated through simulation analysis using human tracked coronary motion data that accounting for this cardiac variability by adaptively changing the trigger delay for acquisition on a beat-to-beat basis improves overall motion compensation and hence MR image quality evaluated in terms of SNR and CNR values.

Accelerating advanced MRI reconstructions on GPUs

Computational acceleration on graphics processing units (GPUs) can make advanced magnetic resonance imaging (MRI) reconstruction algorithms attractive in clinical settings, thereby improving the quality of MR images across a broad spectrum of applications. This paper describes the acceleration of such an algorithm on NVIDIA’s Quadro FX 5600. The reconstruction of a 3D image with 128 3 voxels achieves up to 180 GFLOPS and requires just over one minute on the Quadro, while reconstruction on a quad-core CPU is twenty-one times slower. Furthermore, for the data set studied in this article, the percent error exhibited by the advanced reconstruction is roughly three times lower than the percent error incurred by conventional reconstruction techniques.

SU‐GG‐I‐123: When Things Go Kaboom! Handling the Safety of An MRI Facility During a Nearby Building Implosion

Purpose: To develop methods of assessing the risk associated with implosive demolition of a large building, located in the immediate vicinity of a large MR imaging facility. To devise appropriate protocols for monitoring the impact of building collapse on the status of MR Imaging equipment. Method and Materials: Blasting and demolition industry follows regulations, designed to prevent damage by shock waves to building structures located near the blasting sites. On the other hand, manufacturers of MRI equipment specify limits of mechanical vibration levels allowed at magnet locations. The major problem is, these two approaches focus on very different goals (preserving a building structure vs. ensuring proper quality of MR images). Thus, each party uses different definitions, nomenclature, testing methodology, and compliance guidelines. As a result, statements made by one party (the demolition monitoring experts) are incompatible with concerns raised by the other party (the MRI equipment manufacturers). Information about the blast monitoring technology was reviewed and adopted to the task of assessing the risk of damage to the MRI equipment caused by shockwaves resulting from building collapse. Blast vibration monitoring seismographs have been evaluated to assess sensitivity and precision limits of acquired data. Methods of translating engineering system of units, used by blasting industry, into units used by MRI equipment vendors, have been developed. Results: Two major possible types of damage to MRI equipment have been identified: magnet quenches and structural damage to the RF shielding. Since no prior data regarding recommended limits of blast vibration levels at MRI magnet location were found, a conservative limit of 500 μg was adopted in risk analysis. Conclusion: As a result of careful planning, the building demolition was successful with no demonstrable effect on MRI equipment. The largest recorded vibration level at the MRI site was about 30 μg at 30 Hz.

Dynamic pelvic floor MR imaging at 3 T in patients with clinical signs of urinary incontinence-preliminary results

To prospectively evaluate feasibility, image quality and diagnostic accuracy of dynamic MR imaging the pelvic floor at 3.0 T in patients with urinary incontinence and to compare these results with those of MRI performed at 1.5 T. Ten patients with the diagnosis of urinary incontinence (clinical symptoms, clinical examination, pelvic ultrasound) were examined with a dynamic balanced FFE (B-FFE) sequence at 1.5 T and 3.0 T on the same day in a randomized order. Spatial (1.5 x 1.5 x 8 mm) and temporal (0.44 s) resolution at 3.0 T were comparable to the 1.5-T B-FFE sequence. Two radiologists assessed visual signal to noise (three-point scale), artefact level (five-point scale) and final MR diagnoses with regard to pelvic floor weakness (independent analysis). The diagnoses obtained at 1.5-T field strength and the results of the clinical tests served as standard of reference. In addition, ROI-based quantitative measurements were performed to assess different tissue contrasts at both field strengths. Data were analyzed for statistical differences by using the Wilcoxon's matched pairs test and the marginal homogeneity test. Visual signal to noise was rated higher at 3.0 T for all ten studies by both radiologists. With regard to artefact level, there was no statistically significant difference between the studies obtained at 3.0 T as compared to the corresponding 1.5-T studies (marginal homogeneity test: p = 0.18 for reviewer 1 and 0.41 for reviewer 2). Mean artefact level was rated minor to moderate by both reviewers for both field strengths (excellent interobserver agreement with Kendall-W value of 0.973). Except for a higher tissue contrast between fat and urethra at 1.5 T, there were no statistically significant differences between tissue contrast at 1.5 T as compared to 3.0 T (Wilcoxon's test). Final MR diagnoses regarding pelvic floor weakness did not differ between 3.0-T and 1.5-T field strength and correlated well with the results of the clinical tests. Dynamic pelvic floor MR imaging is feasible at 3.0 T. Our preliminary data indicate that evaluation of pelvic floor disease seems to be possible with 3.0 T equally well as compared to 1.5 T.

Magnetic resonance imaging versus ultrasonography in fetal pathology

This paper describes our experience with magnetic resonance imaging (MRI) in the assessment of fetal anatomical structures and major fetal pathologies. The retrospective study included 128 pregnant women between the 22nd and 38th week of gestation. We used the following imaging protocol: T2-weighted single-shot fast spin-echo sequences for all foetuses and, in selected cases, gradient echo with steady-state free precession (SSFP), T1-weighted spoiled gradient echo [fast low-angle shot (FLASH)] with and without fat saturation, and T2 thick-slab sequences with multiplanar technique. In 32 cases, we performed diffusion-weighted sequences with apparent diffusion coefficient (ACD) maps on the brain, the kidneys and the lungs. We achieved diagnostic-quality images in 125 of 128 patients; MR image quality was unsatisfactory in three cases only. In 16 cases with previous negative ultrasound (US) findings, MRI confirmed the US diagnosis. MRI confirmed the positive US diagnosis in 67 of 109 cases (61.5%); in 11 cases it changed the US diagnosis, and in 31/109 the examination was negative. In addition, MRI identified other anomalies not recognised during US examination. With its ultrafast sequences, fetal MRI provides good detail of normal fetal anatomy and allows characterisation of suspected anomalies.

Basic and Tailored RF Shimming in a Multi-transmit Whole Body MR System

Wave propagation efiects diminish the quality of MR images at main flelds of 3T or above. Parallel RF transmission has the potential of compensating for these efiects through RF shimming. RF shimming can be performed in two difierent ways. The basic way of RF shimming is to adjust the global amplitude and phase of the currents in each transmit element, aiming at a constant B1 in the region of interest. For 3D volume imaging, 3D RF shimming is facilitated using multiple frequencies for the difierent transmit elements. On the other hand, \tailored RF shimming can be performed via multi-dimensional RF pulses designed to achieve a spatially constant excitation pattern. Using an MR system equipped with parallel RF channels, these multi-dimensional RF pulses can be accelerated via Transmit SENSE. The potential of both, basic and tailored RF shimming, can be enhanced signiflcantly, if only a constant B1 amplitude is demanded, and an arbitrary spatial distribution of the resulting B1 phase is allowed. This is the case, if only magnitude images are of interest. However, this approach introduces a non-linear problem, requiring corresponding numerical techniques. Optimal results for basic RF shimming are obtained with transmit arrays owning preferably homogeneous sensitivity distributions of the individual elements. On the other hand, for tailored shimming, the transmit elements require orthogonal sensitivity distributions, which are easier to achieve with inhomogeneous sensitivity distributions. Thus, the transmit coil array cannot be optimized for both, basic and tailored RF shimming simultaneously, and a suitable compromise has to be found. This study compares the difierent RF shimming approaches using a whole-body, 8-channel Tx/Rx system at 3T. It shows, that basic RF shimming is able to remove B1 inhomogeneities to a high degree, both in phantoms and in vivo. Tailored RF shimming is able to achieve even higher levels of B1 homogeneity, which, however, might not be necessary for the majority of clinical applications.

Preoperative 3 T MR imaging of rectal cancer: Local staging accuracy using a two-dimensional and three-dimensional T2-weighted turbo spin echo sequence

Purpose The purpose of our study was to evaluate the image quality and diagnostic performance of two-dimensional (D) turbo spin echo (TSE) and 3D T2-weighted TSE MR imaging in local staging of rectal cancer at 3 T. Materials and methods 3 T phased-array MR imaging was performed in 36 consecutive patients with biopsy-proven rectal cancer. High-resolution 2D TSE images in three planes and 3D TSE images of the rectum were obtained. Two independent observers performed an image quality assessment using eight image quality characteristics. All 2D and 3D datasets were evaluated separately. MR images were prospectively evaluated by two experienced radiologists in consensus with regard to local disease. Total mesorectal excision was used as the standard of reference. The sensitivity, specificity, positive and negative predictive value, and overall accuracy were calculated. Areas under the receiver operating characteristic (ROC) curve (AUC) were determined. Results Twenty-two patients who underwent a total mesorectal excision were enrolled in this study. Significantly more motion artifacts were present with 3D TSE imaging ( P = 0.04). The overall sensitivity, specificity, and accuracy of muscularis propria invasion in rectal cancer using 2D T2-weighted images were 100%, 66%, and 95%, respectively. There was a statistical significant greater AUC using 2D T2-weighted images compared to 3D T2-weighted MR images ( P = 0.04). The ROC curves describing the results of the interpretation of 2D and 3D T2-weighted datasets regarding perirectal tissue invasion showed no statistical significant difference ( P = 0.41). Conclusions In this study, high local staging accuracies with 3 T 2D T2-weighted MR imaging were demonstrated. 3D T2-weighted MR imaging cannot replace 2D MR imaging for local staging of rectal cancer. However, 3D MR imaging can be used for visualization of the complex pelvic anatomy for treatment planning purposes.

Prospective navigator-echo-based real-time triggering of fetal head movement for the reduction of artifacts

The purpose of this study was to evaluate the neuroimaging quality and accuracy of prospective real-time navigator-echo acquisition correction versus untriggered intrauterine magnetic resonance imaging (MRI) techniques. Twenty women in whom fetal motion artifacts compromised the neuroimaging quality of fetal MRI taken during the 28.7 +/- 4 week of pregnancy below diagnostic levels were additionally investigated using a navigator-triggered half-Fourier acquired single-shot turbo-spin echo (HASTE) sequence. Imaging quality was evaluated by two blinded readers applying a rating scale from 1 (not diagnostic) to 5 (excellent). Diagnostic criteria included depiction of the germinal matrix, grey and white matter, CSF, brain stem and cerebellum. Signal-difference-to-noise ratios (SDNRs) in the white matter and germinal zone were quantitatively evaluated. Imaging quality improved in 18/20 patients using the navigator echo technique (2.4 +/- 0.58 vs. 3.65 +/- 0.73 SD, p < 0.01 for all evaluation criteria). In 2/20 patients fetal movement severely impaired image quality in conventional and navigated HASTE. Navigator-echo imaging revealed additional structural brain abnormalities and confirmed diagnosis in 8/20 patients. The accuracy improved from 50% to 90%. Average SDNR increased from 0.7 +/- 7.27 to 19.83 +/- 15.71 (p < 0.01). Navigator-echo-based real-time triggering of fetal head movement is a reliable technique that can deliver diagnostic fetal MR image quality despite vigorous fetal movement.

MRI/linac integration

Purpose/objectives In radiotherapy the healthy tissue involvement still poses serious dose limitations. This results in sub-optimal tumour dose and complications. Daily image guided radiotherapy (IGRT) is the key development in radiation oncology to solve this problem. MRI yields superb soft-tissue visualization and provides several imaging modalities for identification of movements, function and physiology. Integrating MRI functionality with an accelerator can make these capacities available for high precision, real time IGRT. Design and results The system being built at the University Medical Center Utrecht is a 1.5 T MRI scanner, with diagnostic imaging functionality and quality, integrated with a 6 MV radiotherapy accelerator. The realization of a prototype of this hybrid system is a joint effort between the Radiotherapy Department of the University of Utrecht, the Netherlands, Elekta, Crawley, U.K., and Philips Research, Hamburg, Germany. Basically, the design is a 1.5 T Philips Achieva MRI scanner with a Magnex closed bore magnet surrounded by a single energy (6 MV) Elekta accelerator. Monte Carlo simulations are used to investigate the radiation beam properties of the hybrid system, dosimetry equipment and for the construction of patient specific dose deposition kernels in the presence of a magnetic field. The latter are used to evaluate the IMRT capability of the integrated MRI linac. Conclusions A prototype hybrid MRI/linac for on-line MRI guidance of radiotherapy (MRIgRT) is under construction. The aim of the system is to deliver the radiation dose with mm precision based on diagnostic quality MR images.

Construction of a modified capacitive overlap MR coil for imaging of small animals and objects in a clinical whole-body scanner

During the last two decades, there has been an explosive increase in the number of MR investigations involving genetically manipulated mice and rats. Many of the animal studies are performed in a more or less clinical environment, where whole-body MR scanners are the only option available. The quality and acquisition time of MR images have improved with the development of novel RF coil technologies. This communication describes the construction of a small inductively coupled capacitive overlap transmit-receive MR coil for imaging of small animals and objects in a clinical MR scanner. The MR coil presented here is a modified version of the bridged loop-gap coil and consists of two tube-shaped coupled resonance circuits, where the primary circuit partly encapsulates the imaging (secondary) circuit. By rotating the concentric primary coil relative to the secondary coil tuning over a range of several hundred kilohertz is obtained. The coil performance was characterized experimentally by acquiring high-resolution anatomical, diffusion and perfusion MR images as well as the acquisition of proton spectra of a mouse tumour.

Abstract 1454: Targeted Molecular Imaging Probes For In Vivo MR Detection Of Atherosclerotic Lesions Using Antibodies Against Oxidized Low Density Lipoprotein

Background: Oxidized low-density lipoprotein (OxLDL) plays a key role in the initiation, progression and destabilization of atherosclerotic plaque. Molecular imaging probes that target OxLDL may allow in vivo detection of vulnerable plaques. In this study, magnetic resonance imaging (MRI) was used to detect atherosclerotic lesions in apolipoprotein deficient mice (ApoE−/−) using micelles comprised of gadolinium lipids, fluorescent rhodamine, PEG-lipids, and MDA2, a murine monoclonal antibody that binds malondialdehyde (MDA) lysine epitopes present in OxLDL. Materials and Results: Untargeted micelles, MDA2-labeled micelles and nonspecific polyclonal IgG micelles were prepared and characterized with respect to OxLDL binding capacity, pharmacokinetics, and biodistribution in wild type (WT) and ApoE−/− mice. MR imaging was performed at 9.4T over a 3-week interval after administration of 0.075 mmol Gd/kg micelles. MDA2 increased the micelle size, blood half-life, and MR efficacy relative to untargeted and IgG-micelles. Maximal plaque enhancement (&gt;125%) was observed 72 hours post MDA2-micelle injection (Figure ). Untargeted and IgG-micelles did not exhibit significant wall enhancement at any of the time points studied. Confocal microscopy revealed that MDA2-micelles accumulate within foam cells associated with atherosclerotic plaque. WT mice showed no significant MR wall enhancement for any of the micelles studied. Conclusions: MR imaging using MDA2-micelles demonstrates specific targeting of OxLDL and foam cells and provides excellent MR image quality. This study suggests that it may be feasible to image similar atherosclerotic lesions in humans with MRI.

Magnetic susceptibility and electrical conductivity of metallic dental materials and their impact on MR imaging artifacts

Objectives The aim of this study was to test the hypothesis that dental materials vary significantly in MR-relevant material parameters—magnetic susceptibility and electrical conductivity, and that knowledge of these parameters may be used to estimate the quality of MR imaging in the presence of devices made of such materials. Methods Magnetic susceptibility, electrical conductivity and artifacts were evaluated for 45 standardized cylindrical samples of dental alloys and amalgams. Magnetic susceptibility was determined by fitting the phase of gradient-echo MR images to numerically modeled data. Electrical conductivity was determined by standard electrotechnical measurements. Artifact sizes were measured in spin-echo (SE) and gradient-echo (GE) images at 1.5 T according to the standards of the American Society for Testing and Materials. Results It has been confirmed that dental materials differ considerably in their magnetic susceptibility, electrical conductivity and artifacts. For typical dental devices, magnetic susceptibility differences were found of little clinical importance for diagnostic SE/GE imaging of the neck and brain, but significant for orofacial imaging. Short-TE GE imaging has been found possible even in very close distances from dental devices made of amalgams, precious alloys and titanium alloys. Nickel–chromium and cobalt–chromium artifacts were found still acceptable, but large restorations of aluminum bronzes may preclude imaging of the orofacial region. The influence of electrical conductivity on the artifact size was found negligible. Significance MR imaging is possible even close to dental devices if they are made of dental materials with low magnetic susceptibility. Not all materials in current use meet this requirement.

Effects of simultaneous EEG recording on MRI data quality at 1.5, 3 and 7 tesla

Although the focus of attention on data degradation during simultaneous MRI/EEG recording has to date largely been upon EEG artefacts, the presence of the conducting wires and electrodes of the EEG recording system also causes some degradation of MRI data quality. This may result from magnetic susceptibility effects which lead to signal drop-out and image distortion, as well as the perturbation of the radiofrequency fields, which can cause local signal changes and a global reduction in the signal to noise ratio (SNR) of magnetic resonance images. Here, we quantify the effect of commercially available 32 and 64 electrode caps on the quality of MR images obtained in scanners operating at magnetic fields of 1.5, 3 and 7 T, via the use of MR-based, field-mapping techniques and analysis of the SNR in echo planar image time series. The electrodes are shown to be the dominant source of magnetic field inhomogeneity, although the localised nature of the field perturbation that they produce means that the effect on the signal intensity from the brain is not significant. In the particular EEG caps investigated here, RF inhomogeneity linked to the longer ECG and EOG leads causes some reduction in the signal intensity in images obtained at 3 and 7 T. Measurements of the standard deviation of white matter signal in EPI time series indicates that the introduction of the EEG cap produces a small reduction in the image signal to noise ratio, which increases with the number of electrodes used.

SU-FF-J-121: MRI of Nasopharygneal Carcinoma (NPC) Patients in Radiation Treatment Position to Improve the Delineation of the Tumour and Cervical Lymph Nodes

Purpose: The existing MR imaging technique for NPC patients precludes the use of any immobilization device and inclusion of adequate neck coverage due to the design of the head coil used. We established an MR imaging protocol for NPC patients performed in the radiation treatment position to facilitate the MR/CT image registration and to include adequate coverage to the lower neck. The goal is to reduce the uncertainties associated with the localization of the tumour and cervical lymph nodes. Method and Materials: An MR compatible base‐plate which accommodates the immobilization mask was fabricated. We employed two large surface coils (one anterior and one posterior) to acquire the MRI with the patient immobilized in a thermoplastic mask in the radiation treatment position. Seven NPC patients were selected and, for each patient, two series of MR images, namely T1‐weighted with contrast and T2‐weighted, were taken respectively. Both series included images from the vertex to the lower neck. The MR images were then registered with the planning CTimages by an automatic image registration software for contouring purposes. Results: The same orientation for the MR and CTimages eliminates the necessity of reformatting the MR images to a plane as that of the CTimages during the registration. This reduces the degradation of image quality no matter what algorithm of image registration is used. The inclusion of the lower neck in the MR images greatly improves the delineation of the cervical lymph nodes. Conclusion: The present method provides MR images in a radiation treatment position with adequate lower neck coverage. The MR image quality is good enough for contouring purposes. The accurate registered MR images are particularly valuable in reducing the uncertainties associated with the localization of the tumour especially near the skull base and of the cervical lymph nodes.

Development and implementation of an MR-compatible whole body video system

In conventional fMRI, the possibilities for remote visual monitoring of a volunteer or patient during scanning are very limited. However, performance during motor paradigms, monitoring of ventilated patients, or patients with movement disorders is often of particular interest. In order to facilitate remote visual monitoring during MRI examinations, an MR-compatible video system was developed and implemented to allow monitoring of the face and the body, separately, with two cameras. The system reliably allows the recording of video data during MRI scanning without affecting MR image quality. The applicability of the system for the online recording of tics during an fMRI study with Tourette patients is demonstrated. Monitoring of a motor task in healthy controls and Tourette patients was possible with exact synchronisation to the fMRI paradigm. Details of system set up and example control measurements are presented. Further, electronic and mechanical components are described in detail enabling easy reproduction of the system.

Magnetic field perturbation of neural recording and stimulating microelectrodes

To improve the overall temporal and spatial resolution of brain mapping techniques, in animal models, some attempts have been reported to join electrophysiological methods with functional magnetic resonance imaging (fMRI). However, little attention has been paid to the image artefacts produced by the microelectrodes that compromise the anatomical or functional information of those studies. This work presents a group of simulations and MR images that show the limitations of wire microelectrodes and the potential advantages of silicon technology, in terms of image quality, in MRI environments. Magnetic field perturbations are calculated using a Fourier-based method for platinum (Pt) and tungsten (W) microwires as well as two different silicon technologies. We conclude that image artefacts produced by microelectrodes are highly dependent not only on the magnetic susceptibility of the materials used but also on the size, shape and orientation of the electrodes with respect to the main magnetic field. In addition silicon microelectrodes present better MRI characteristics than metallic microelectrodes. However, metallization layers added to silicon materials can adversely affect the quality of MR images. Therefore only those silicon microelectrodes that minimize the amount of metallic material can be considered MR-compatible and therefore suitable for possible simultaneous fMRI and electrophysiological studies. High resolution gradient echo images acquired at 2 T (TR/TE = 100/15 ms, voxel size = 100 × 100 × 100 µm3) of platinum–iridium (Pt–Ir, 90%–10%) and tungsten microwires show a complete signal loss that covers a volume significantly larger than the actual volume occupied by the microelectrodes: roughly 400 times larger for Pt–Ir and 180 for W, at the tip of the microelectrodes. Similar MR images of a single-shank silicon microelectrode only produce a partial volume effect on the voxels occupied by the probe with less than 50% of signal loss.

Comparison and evaluation of mouse cardiac MRI acquired with open birdcage, single loop surface and volume birdcage coils

Although the quality and speed of MR images have vastly improved with the development of novel RF coil technologies, the engineering expertise required to implement them is often not available in many animal in vivo MR laboratories. We present here an open birdcage coil design which is easily constructed with basic RF coil expertise and produces high quality images. The quality and advantages of mouse cardiac MR images acquired with open birdcage coils were evaluated and compared to images acquired with a bent single loop surface, and standard birdcage coils acquired at 4.7 Tesla. Two low pass open birdcage coils, two single loop surface coils, and a low pass volume birdcage coil were constructed and their B1 distributions were evaluated and compared. The calculated average signal-to-noise ratio for the left ventricular wall was 10, 23 and 32 for the volume birdcage coil, single loop surface coil and open birdcage coil, respectively. The results demonstrate that the open birdcage coil provides greater sensitivity than the volume coil and a higher signal/contrast-to-noise ratio and B1 homogeneity than the single loop surface coil. The open birdcage coil offers easy access and better quality mouse cardiac imaging than both the single loop surface coil and volume birdcage coil and does not require extensive RF engineering expertise to construct.

MR imaging of the cervical spine: assessment of image quality with parallel imaging compared to non-accelerated MR measurements

To compare the quality of cervical spine MR images obtained by parallel imaging [generalized autocalibrating partially parallel acquisition (GRAPPA)] with those of non-accelerated imaging, we conducted both phantom studies and examinations of ten volunteers at 1.5Tesla with a dedicated 12-element coil system and a head-spine-neck coil combination. Acquisitions included axial T2-weighted (T2w) images with both methods, and sagittal T2w and T1w images in vivo with the latter coil combination. Non-accelerated MRI with two averages and GRAPPA (acceleration factor 2) with two averages (GRAPPA/2AV, time reduction of approximately 50%) and four averages (GRAPPA/4AV) were compared. In the phantom, the signal-to-noise ratio of the GRAPPA/2AV was lower than that of the other two settings. In vivo, the image inhomogeneity (non-uniformity, NU) was significantly higher in T2w GRAPPA/2AV than in both other settings, and in T1w GRAPPA/2AV compared to GRAPPA/4AV. Subjectively, the delineation of anatomical structures was sufficient in all sequences. Only the spinal cord was considered to be better delineable on the non-accelerated T1w sequence compared to GRAPPA/2AV. In part, GRAPPA/4AV performed better than the other settings. The subjective image noise was lowest with GRAPPA/4AV. In cervical spine MRI, the examination time can be reduced by nearly 42% with GRAPPA, while preserving sufficient image quality.

SU‐FF‐I‐69: The Impact of Sparse Data Sets On the Quality of MR Images

Purpose: To evaluate the impact of reduction in acquired data size on the final appearance of MR images and to assess the synergetic impact of image interpolation, used in softcopy viewing, on the perception of image quality. Method and Materials: A rapid progress in signal quality of the MRI data acquired with modern MRI hardware, coupled with a desire to speed up the acquisition times as much as possible, promotes a practice where drastic reductions in the size of acquired data are implemented by the MRI system operator. Routinely, the physician viewing the MR images is presented with incomplete and confusing information about the true size of the acquired data. Most physicians are oblivious to the potential impact of complex acquisition algorithms on the quality of images. The situation is further confounded by various image interpolation techniques used by softcopy display systems.A standard spin echo imaging sequence (270/10, FOV 30 cm, slice thickness 5 mm, gap 5 mm, 11 slices, 1 NEX, BW=2*15.63 kHz) was used to generate a baseline image with 256*256 acquisition matrix. This was followed by an acquisition of a 512*512 image that served as a “golden reference standard”, given the fact that the standard MR displays use 512*512 image size as a default. The baseline protocol was then repeated by varying options and parameters that affected the sizes of the acquisition matrix, processing matrix, or display matrix. Results: Images other than the golden reference standard exhibit artifacts whose strength strongly depends on the resampling algorithm applied. The extent and severity of observed artifacts increases with the number of options used to generate the image. Conclusion: The image acquisition parameters and display resampling algorithms need to be carefully accounted for when evaluating image artifacts in MR images.