Quiet MR sequences in clinical routine: initial experience in abdominal imaging.

OBJECTIVE: Purpose of our study was to demonstrate the feasibility and limitations of acoustic noise reduction in a standard clinical MRI protocol for abdominal imaging. METHODS: Acoustic noise and image quality were assessed in 17 patients for a standard liver imaging protocol including TSE and GRE sequences and compared to quiet optimizations as described by Heismann et al. Two blinded readers scored artifacts, the delineation of the abdominal organs and level of diagnostic confidence. Means of the sound level measurements, the ratings and the measurement of SNR and CNR were compared. RESULTS: Significant reduction of acoustic noise was measured for T2 TSE (-30.7%), T2 HASTE (-22.6%) and less difference for T1 DIXON (-4.7%) and T1 FLASH (-2.7%). SNR and CNR were lower for quiet T2 TSE (-18.0%, -23.1%) and T2 HASTE (-46.2%, -37.7%) and higher for T1 DIXON (+32.0%, +24.4%). Inter-rater correlation was k = 0.987 with p < 0.001. CONCLUSIONS: Although sequence-based noise optimizations faces problems in T1 FLASH and DIXON sequences, there is an important acoustic benefit in T2 TSE and T2 HASTE sequences, which goes along with a maintained image quality and diagnostic confidence.

Arterial spin labeled carotid MR angiography: A phantom study examining the impact of technical and hemodynamic factors.

PURPOSE: To quantify the accuracy of three-dimensional (3D) radial arterial spin labeled (ASL) magnetic resonance angiography (MRA) using vascular models of carotid stenosis. METHODS: Eight vascular models were imaged at 1.5 Tesla using pulsatile flow waveforms at rates found in the internal carotid arteries (100-400 mL/min). The impacts of the 3D ASL imaging readout (fast low angle shot (FLASH) versus balanced steady-state free precession (bSSFP)), ultrashort echo time imaging using a pointwise encoding time reduction with radial acquisition (PETRA), and model stenosis severity on the accuracy of vascular model display at the location of stenosis were quantified. Accuracy was computed vis-à-vis a reference bSSFP volume acquired under no flow. Comparisons were made with standard-of-care contrast-enhanced MRA (CEMRA) and Cartesian time-of-flight (TOF) MRA protocols. RESULTS: For 50% and 70% stenoses, CEMRA was most accurate (respective accuracies of 81.7% and 78.6%), followed by ASL FLASH (75.7% and 71.8%), ASL PETRA (69.6% and 70.6%), 3D TOF (66.6% and 57.1%), ASL bSSFP (68.7% and 51.2%), and 2D TOF (65.1% and 50.6%). CONCLUSION: Flow phantom imaging studies show that ASL MRA can improve the display of hemodynamically significant carotid arterial stenosis compared with TOF MRA, with FLASH and ultrashort echo time readouts being most accurate.

Quiet diffusion-weighted head scanning: Initial clinical evaluation in ischemic stroke patients at 1.5T.

PURPOSE: To compare the quality and diagnostic value of routine single-shot, echo-planar imaging, diffusion-weighted imaging (ss-EPI-DWI) to those of quiet readout segmented EPI-DWI (q-DWI) in magnetic resonance imaging (MRI) of acute stroke. MATERIALS AND METHODS: Twenty-six patients with acute stroke underwent a 1.5T MRI including diffusion-weighted ss-EPI and q-DWI. The two sequences were protocolled to have identical spatial resolution and spatial coverage. q-DWI was tested with (regular q-DWI) and without (fast q-DWI) averaging in 13 patients each. The acoustic noise generated by each sequence was measured. Quantitative and qualitative assessments regarding signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), lesion conspicuity, level of artifacts, overall image quality as well as diagnostic content were performed. RESULTS: SNR and CNR values of the q-DWI scans were considerably higher than those of ss-EPI DWI (P ≤ 0.0078). No statistical difference was found for lesion conspicuity (P ≥ 0.125). Statistical differences were found for level of artifacts (P ≥ 0.0078) and overall image quality (P ≥ 0.002). Both were evaluated better in the ss-EPI DWI than in the regular and fast q-DWI. Apart from one fast q-DWI patient, radiologists voted the images to have the same diagnostic content, with upper 90% confidence limits of 0.238 for regular q-DWI and 0.429 for fast q-DWI. CONCLUSION: If the acoustic burden is critical to the patient, q-DWI is an equivalent quiet alternative to ss-EPI DWI for use in stroke patients. J. Magn. Reson. Imaging 2016;44:1238-1243.

MR neurographic orthopantomogram: Ultrashort echo-time imaging of mandibular bone and teeth complemented with high-resolution morphological and functional MR neurography.

PURPOSE: Panoramical radiographs or cone-beam computed tomography (CT) are the standard-of-care in dental imaging to assess teeth, mandible, and mandibular canal pathologies, but do not allow assessment of the inferior alveolar nerve itself nor of its branches. We propose a new technique for "MR neurographic orthopantomograms" exploiting ultrashort echo-time (UTE) imaging of bone and teeth complemented with high-resolution morphological and functional MR neurography. MATERIALS AND METHODS: The Institutional Review Board approved the study in 10 healthy volunteers. Imaging of the subjects mandibles at 3.0T (Magnetom Skyra, Siemens-Healthcare) using a 64-channel head coil with isotropic spatial resolution for subsequent multiplanar reformatting, was performed. Bone images were acquired using a 3D PETRA sequence (TE, 0.07 msec). Morphological nerve imaging was performed using a dedicated 3D PSIF and 3D SPACE STIR sequence. Functional MR neurography was accomplished using a new accelerated diffusion-tensor-imaging (DTI) prototype sequence (2D SMS-accelerated RESOLVE). Qualitative and quantitative image analysis was performed and descriptive statistics are provided. RESULTS: Image acquisition and subsequent postprocessing into the MR neurographic orthopantomogram by overlay of morphological and functional images were feasible in all 10 volunteers without artifacts. All mandibular bones and mandibular nerves were assessable and considered normal. Fiber tractography with quantitative evaluation of physiological diffusion properties of mandibular nerves yielded the following mean ± SD values: fractional anisotropy, 0.43 ± 0.07; mean diffusivity (mm(2) /s), 0.0014 ± 0.0002; axial diffusivity, 0.0020 ± 0.0002, and radial diffusivity, 0.0011 ± 0.0001. CONCLUSION: The proposed technique of MR neurographic orthopantomogram exploiting UTE imaging complemented with high-resolution morphological and functional MR neurography was feasible and allowed comprehensive assessment of osseous texture and neural microarchitecture in a single examination. J. Magn. Reson. Imaging 2016;44:393-400.

Quiet T1-weighted imaging using PETRA: initial clinical evaluation in intracranial tumor patients.

PURPOSE: To compare the lesion contrast and signal to noise ratio (SNR) obtained with T1-weighted pointwise encoding time reduction with radial acquisition (PETRA) to those of Magnetization-Prepared RApid Gradient-Echo (MPRAGE) for contrast-enhanced imaging of primary and metastatic intracranial tumors, and to investigate whether PETRA is able to reduce acoustic noise for improved patient comfort. MATERIALS AND METHODS: Fifteen patients with intracranial tumors underwent 3 Tesla MRI including inversion-prepared PETRA and MPRAGE. The two sequences had comparable scan times, spatial resolution and spatial coverage. "Tumor conspicuity" was rated qualitatively by two radiologists, while enhancing lesion-to-white matter contrast to noise ratio (CNR) and white-matter SNR were analyzed quantitatively using paired t-tests. The acoustic noise generated by each sequence was measured. RESULTS: Qualitative rating of "tumor conspicuity" by two radiologists resulted in nearly identical average scores for the two sequences. Quantitative analyses revealed that (i) there was no significant difference between the mean CNR values of the two sequences (P = 0.57), (ii) the mean SNR of PETRA was significantly higher than that of MPRAGE (P < 0.01), and (iii) the mean sound level of PETRA was significantly lower than that of MPRAGE (P < 0.01). CONCLUSION: Inversion-prepared PETRA was found to be viable as a quiet alternative to MPRAGE for contrast-enhanced T1-weighted studies of intracranial tumors.

Acoustic-noise-optimized diffusion-weighted imaging.

OBJECTIVE: This work was aimed at reducing acoustic noise in diffusion-weighted MR imaging (DWI) that might reach acoustic noise levels of over 100 dB(A) in clinical practice. MATERIALS AND METHODS: A diffusion-weighted readout-segmented echo-planar imaging (EPI) sequence was optimized for acoustic noise by utilizing small readout segment widths to obtain low gradient slew rates and amplitudes instead of faster k-space coverage. In addition, all other gradients were optimized for low slew rates. Volunteer and patient imaging experiments were conducted to demonstrate the feasibility of the method. Acoustic noise measurements were performed and analyzed for four different DWI measurement protocols at 1.5T and 3T. RESULTS: An acoustic noise reduction of up to 20 dB(A) was achieved, which corresponds to a fourfold reduction in acoustic perception. The image quality was preserved at the level of a standard single-shot (ss)-EPI sequence, with a 27-54% increase in scan time. CONCLUSIONS: The diffusion-weighted imaging technique proposed in this study allowed a substantial reduction in the level of acoustic noise compared to standard single-shot diffusion-weighted EPI. This is expected to afford considerably more patient comfort, but a larger study would be necessary to fully characterize the subjective changes in patient experience.

Ultrashort echo time imaging using pointwise encoding time reduction with radial acquisition (PETRA).

Sequences with ultrashort echo times enable new applications of MRI, including bone, tendon, ligament, and dental imaging. In this article, a sequence is presented that achieves the shortest possible encoding time for each k-space point, limited by pulse length, hardware switching times, and gradient performance of the scanner. In pointwise encoding time reduction with radial acquisition (PETRA), outer k-space is filled with radial half-projections, whereas the centre is measured single pointwise on a Cartesian trajectory. This hybrid sequence combines the features of single point imaging with radial projection imaging. No hardware changes are required. Using this method, 3D images with an isotropic resolution of 1 mm can be obtained in less than 3 minutes. The differences between PETRA and the ultrashort echo time (UTE) sequence are evaluated by simulation and phantom measurements. Advantages of pointwise encoding time reduction with radial acquisition are shown for tissue with a T(2) below 1 ms. The signal to noise ratio and Contrast-to-noise ratio (CNR) performance, as well as possible limitations of the approach, are investigated. In-vivo head, knee, ankle, and wrist examples are presented to prove the feasibility of the sequence. In summary, fast imaging with ultrashort echo time is enabled by PETRA and may help to establish new routine clinical applications of ultrashort echo time sequences.

High-performance low field MRI enables visualization of persistent pulmonary damage after COVID-19.

The outbreak of coronavirus disease 2019 (COVID-19) with the origin of the spread assumed to be located in Wuhan, China, began in December 2019, and is continuing until now. With the COVID-19 pandemic showing a progressive spread throughout the countries of the world, there is emerging interest for the potential long-term consequences of suffering from a COVID-19 pneumonia. Imaging plays a central role in the diagnosis and management of COVID-19 pneumonia, with chest X-ray examinations and computed tomography (CT) being undoubtedly the modalities most widely used, allowing for a fast and sensitive detection of infiltration patterns associated with COVID-19 pneumonia. For a better understanding of underlying pathomechanisms of pulmonary damage, longitudinal imaging series are warranted, for which CT is of limited usability due to repeated exposure of X-rays. Recent advances in MRI suggested that high-performance low-field MRI might represent a valuable method for pulmonary imaging without the need of radiation exposure. However, so far, low-field MRI has not been applied to study pulmonary damage after COVID-19 pneumonia. We present a case report of a patient who suffered from COVID-19 pneumonia using 0.55 T MRI for follow-up examinations three months after initial infection. Low-field MRI enables a precise visualization of persistent pulmonary changes including ground-glass opacities, which are consistent with CT performed on the same day. Low-field MRI seems to be feasible in the detection of pulmonary involvement in patients with COVID-19 pneumonia and may have the potential for repetitive lung examinations in monitoring the reconvalescence after pulmonary infections.

Quiet FLAIR at 7T MRI.

OBJECTIVE: The aim of this study was to investigate acoustic noise reduction and image quality of cranial magnetic resonance imaging (MRI) at 7T MRI with and without sequence-based acoustic noise reduction. MATERIALS AND METHODS: Fifteen patients and 5 healthy volunteers underwent 7T MRI scanning. A fluid-attenuated inversion recovery (FLAIR) sequence was acquired with and without sequence-based acoustic noise reduction. The acoustic noise generated by each sequence was measured. Quantitative and qualitative assessments regarding signal-to-noise ratio, contrast-to-noise ratio, lesion conspicuity, level of artifacts, and overall image quality were performed. Furthermore, detection rates of white matter lesions were evaluated by 2 observers for both sequences. RESULTS: Acoustic noise was significantly reduced from initially 92.7 dB(A) to 78.9 dB(A), corresponding to an 80% reduction in sound pressure. The visual assessment revealed no significant difference in the level of artifacts. Although rated very high by both readers, lesion conspicuity and image quality averaged better for the regular FLAIR sequence. Signal-to-noise ratio and contrast-to-noise ratio slightly decreased when applying the sequence-based acoustic noise reduction. No significant difference was found between the detection rates of white matter lesions for both observers. CONCLUSIONS: Reducing sound pressure by 80% in FLAIR imaging at 7T ultra-high-field MRI is feasible while maintaining high diagnostic image quality.

Morphological MR imaging of the articular cartilage of the knee at 3 T-comparison of standard and novel 3D sequences.

OBJECTIVES: This study sought to compare various 3D cartilage sequences and to evaluate the usefulness of ultrashort TE (UTE) imaging, a new technique to isolate signal from the osteochondral junction. METHODS: Twenty knees were examined at 3 T with 3D spoiled GRE (FLASH), double-echo steady-state (DESS), balanced SSFP, 3D turbo spin-echo (TSE), and a prototype UTE sequence. Two radiologists independently evaluated all images. Consensus readings of all sequences were the reference standard. Statistical analysis included Friedman and pairwise Wilcoxon tests. Retrospective analysis of UTE morphology of osteochondral tissue in normal and abnormal cartilage seen at conventional MR was also performed. RESULTS: Three-dimensional TSE was superior to other sequences for detecting cartilage lesions. FLASH and DESS performed best in the subjective quality analysis. On UTE images, normal cartilage exhibited a high-intensity linear signal near the osteochondral junction. Retrospective analysis revealed abnormal UTE morphology of the osteochondral junction in 50 % of cartilage lesions diagnosed at conventional MR. CONCLUSIONS: Cartilage imaging of the knee at 3 T can be reliably performed using 3D TSE, showing high accuracy when compared to standard sequences. Although UTE depicts signal from the deep cartilage layer, further studies are needed to establish its role for assessment of cartilage. MAIN MESSAGES: • MRI is the best available imaging method for assessment of knee cartilage. • Cartilage imaging can be reliably performed using 3D TSE. • UTE cannot be used as a single sequence to assess cartilage.

Quantitative characterizations of ultrashort echo (UTE) images for supporting air-bone separation in the head.

Accurate separation of air and bone is critical for creating synthetic CT from MRI to support Radiation Oncology workflow. This study compares two different ultrashort echo-time sequences in the separation of air from bone, and evaluates post-processing methods that correct intensity nonuniformity of images and account for intensity gradients at tissue boundaries to improve this discriminatory power. CT and MRI scans were acquired on 12 patients under an institution review board-approved prospective protocol. The two MRI sequences tested were ultra-short TE imaging using 3D radial acquisition (UTE), and using pointwise encoding time reduction with radial acquisition (PETRA). Gradient nonlinearity correction was applied to both MR image volumes after acquisition. MRI intensity nonuniformity was corrected by vendor-provided normalization methods, and then further corrected using the N4itk algorithm. To overcome the intensity-gradient at air-tissue boundaries, spatial dilations, from 0 to 4 mm, were applied to threshold-defined air regions from MR images. Receiver operating characteristic (ROC) analyses, by comparing predicted (defined by MR images) versus 'true' regions of air and bone (defined by CT images), were performed with and without residual bias field correction and local spatial expansion. The post-processing corrections increased the areas under the ROC curves (AUC) from 0.944 ± 0.012 to 0.976 ± 0.003 for UTE images, and from 0.850 ± 0.022 to 0.887 ± 0.012 for PETRA images, compared to without corrections. When expanding the threshold-defined air volumes, as expected, sensitivity of air identification decreased with an increase in specificity of bone discrimination, but in a non-linear fashion. A 1 mm air mask expansion yielded AUC increases of 1 and 4% for UTE and PETRA images, respectively. UTE images had significantly greater discriminatory power in separating air from bone than PETRA images. Post-processing strategies improved the discriminatory power of air from bone for both UTE and PETRA images, and reduced the difference between the two imaging sequences. Both post-processed UTE and PETRA images demonstrated sufficient power to discriminate air from bone to support synthetic CT generation from MRI data.

Correcting slice selectivity in hard pulse sequences.

Many MRI sequences use non-selective hard pulse excitation in the presence of imaging gradients. In this work, we investigate to which extent the sinc-shaped frequency excitation profiles of the pulse can be used for imaging without the generation of artefacts. A correction algorithm is proposed that eliminates the influence of the excitation profile. Phantom as well as in vivo measurements prove that enhanced image quality can be obtained as long as the first minimum of the excitation profile lies outside the imaged object.