Establishing an objective biomarker for corneal cystinosis using a threshold-based Spectral domain optical coherence tomography imaging algorithm

PURPOSE

The purpose of the present study was to establish a semi-automated threshold-based image segmentation algorithm to detect and objectively quantify corneal cystine crystal deposition in ocular cystinosis with anterior segment optical coherence tomography (AS-OCT).

METHODS

This prospective, observational, comparative study included 88 eyes of 45 patients from the German Cystinosis Registry Study as well as 68 eyes of 35 healthy control subjects. All eyes were imaged with AS-OCT (Cirrus HD-OCT 5000, Carl Zeiss Meditec AG, Jena, Germany). As an initial step, B-scan images were subjectively analysed for typical changes in morphology in comparison to healthy controls. Based on the experience gained, an objective semi-automated B-scan image segmentation algorithm was developed using a grey scale value-based threshold method to automatically quantify corneal crystals.

RESULTS

On AS-OCT B-scans, corneal crystals appeared as hyperreflective deposits within the corneal stroma. The crystals were distributed either in all stromal layers (43 eyes, 49%) or confined to the anterior (23 eyes, 26%) or posterior stroma (22 eyes, 25%), respectively. The novel automatic B-scan image segmentation algorithm was most efficient in delineating corneal crystals at higher grey scale thresholds (e.g. 226 of a maximum of 255). Significant differences in suprathreshold grey scale pixels were observable between cystinosis patients and healthy controls (p < 0.001). In addition, the algorithm was able to detect an age-dependent depth distribution profile of crystal deposition.

CONCLUSION

Objective quantification of corneal cystine crystal deposition is feasible with AS-OCT and can serve as a novel biomarker for ocular disease control and topical treatment monitoring.

A new concept for improved quantitative analysis of reversible transverse relaxation in tissues with variable microscopic field distribution

PURPOSE

The intravoxel distribution of the magnetic field strongly influences signal dephasing after RF excitation and the resulting signal decay in gradient echo-based MRI. In this work, several different field distribution models were applied and tested for analysis of microscopic field characteristics within pixels.

THEORY

A flexible model for improved pixel-wise characterization of the underlying field distribution is introduced. The proposed symmetric alpha-stable (SαS) distribution covers Lorentzian, Gaussian, and intermediate field distributions in a continuous way using a two-parametric (width and shape) function.

METHODS

The new model was applied on human brain, potatoes (homogeneous isotropic tissue), and stems of pineapple (anisotropic fibrous tissue). Effects of microscopic structure and background gradients on the shape and the widths of the microscopic field distribution were analyzed using gradient echo sampling of the spin echo and multigradient-echo sequences. Effects of non-Lorentzian shapes of microscopic field distributions on the results of common T 2 ∗ measurements with mono-exponential fitting of signal values were tested.

RESULTS

Many pixels of the examined objects showed field characteristics in between Lorentzian and Gaussian shapes. Microscopic field inhomogeneities caused by microscopic susceptibility effects and background gradients sometimes led to rather Gaussian than Lorentzian field distribution. In cases with nearly Gaussian field distribution, mono-exponential fitting of the signal decay resulted in different T 2 ∗ values, depending on the sampling points.

CONCLUSIONS

Using the concept of more flexible distributions for characterization of microscopic susceptibility effects in tissue provides better fitting of data and nearly sampling point-independent results than common T 2 ∗ measurements with mono-exponential fitting.

Volumetric assessment of spontaneous mechanical activities by simultaneous multi-slice MRI techniques with correlation to muscle fiber orientation

The purpose of this work was assessment of volumetric characteristics of spontaneous mechanical activities in musculature (SMAMs) by diffusion-weighted simultaneous multi-slice (DW-SMS) imaging and spatial correlation to anatomical structure, as revealed by fusion to fiber tractographic information derived from diffusion-tensor imaging (DTI). The feasibility of using DW-SMS to image spontaneous events in human musculature was assessed by phantom measurements. Series of DW-SMS images and DTI datasets were recorded from the resting calf of three human subjects. Simultaneously recorded SMAMs in multiple slices were analyzed regarding spatial extension by the Kolmogorov-Smirnov test. Direct correlation of spatial distribution of SMAMs and fiber orientation was investigated by mapping of muscle fibers to multi-slice SMAM datasets. The DW-SMS strategy allows simultaneous assessment of SMAMs in several slices of resting skeletal musculature, since 73.9% of SMAM-affected volumes have shown SMAMs in multiple DW-SMS slices. Spatial extension of SMAMs was highly correlated over different simultaneously recorded DW-SMS slices, and affected areas followed the orientation of muscle fibers with a connectivity ratio up to 57.18 ± 14.80% based on event count and connectivity count maps. In 89.2% of all SMAM-affected datasets muscle fiber connectivity was shown in at least two adjacent slices. Direct correlation between SMAMs in human lower leg musculature and underlying anatomical structure was revealed by high muscle fiber connectivity (89.2%). SMAMs have shown a wide distribution along the longitudinal muscle direction (73.9% in multiple DW-SMS slices) with direct involvement of muscle fibers. Correlation between SMAMs in multiple DW-SMS slices and crossing muscular fiber tracts provides evidence that SMAMs result from physiological processes in musculature. Fusion of DW-SMS with DTI facilitates non-invasive studies of muscle fiber involvement in SMAMs in resting muscle.

An Efficient Modelling-Simulation-Analysis Workflow to Investigate Stump-Socket Interaction Using Patient-Specific, Three-Dimensional, Continuum-Mechanical, Finite Element Residual Limb Models

The lack of an efficient modelling-simulation-analysis workflow for creating and utilising detailed subject-specific computational models is one of the key reasons why simulation-based approaches for analysing socket-stump interaction have not yet been successfully established. Herein, we propose a novel and efficient modelling-simulation-analysis workflow that uses commercial software for generating a detailed subject-specific, three-dimensional finite element model of an entire residual limb from Diffusion Tensor MRI images in <20 min. Moreover, to complete the modelling-simulation-analysis workflow, the generated subject-specific residual limb model is used within an implicit dynamic FE simulation of bipedal stance to predict the potential sites of deep tissue injury. For this purpose, a nonlinear hyperelastic, transversely isotropic skeletal muscle constitutive law containing a deep tissue injury model was implemented in LS-DYNA. To demonstrate the feasibility of the entire modelling-simulation-analysis workflow and the fact that detailed, anatomically realistic, multi-muscle models are superior to state-of-the-art, fused-muscle models, an implicit dynamic FE analysis of 2-h bipedal stance is carried out. By analysing the potential volume of damaged muscle tissue after donning an optimally-fitted and a misfitted socket, i.e., a socket whose volume was isotropically shrunk by 10%, we were able to highlight the differences between the detailed individual- and fused-muscle models. The results of the bipedal stance simulation showed that peak stresses in the fused-muscle model were four times lower when compared to the multi-muscle model. The peak interface stress in the individual-muscle model, at the end of bipedal stance analysis, was 2.63 times lower than that in the deep tissues of the stump. At the end of the bipedal stance analysis using the misfitted socket, the fused-muscle model predicted that 7.65% of the residual limb volume was injured, while the detailed-model predicted 16.03%. The proposed approach is not only limited to modelling residual limbs but also has applications in predicting the impact of plastic surgery, for detailed forward-dynamics simulations of normal musculoskeletal systems.

Spontaneous mechanical and electrical activities of human calf musculature at rest assessed by repetitive single-shot diffusion-weighted MRI and simultaneous surface electromyography

PURPOSE

Assessment of temporal and spatial relations between spontaneous mechanical activities in musculature (SMAM) at rest as revealed by diffusion-weighted imaging (DWI) and electrical muscular activities in surface EMG (sEMG). Potential influences of static and radiofrequency magnetic fields on muscular activity on sEMG measurements at rest were examined systematically.

METHODS

Series of diffusion-weighted stimulated echo planar imaging were recorded with concurrent sEMG measurements. Electrical activities in sEMG were analyzed by non-parametric Friedman and two-sample Kolmogorov-Smirnov test. Direct correlation of both modalities was investigated by temporal mapping of electrical activity in sEMG to DWI repetition interval.

RESULTS

Electrical activities in sEMG and number of visible SMAMs in DWI showed a strong correlation (ρ = 0.9718). High accordance between sEMG activities and visible SMAMs in DWI in a near-surface region around sEMG electrodes was achieved. Characteristics of sEMG activities were almost similar under varying magnetic field conditions.

CONCLUSION

Visible SMAMs in DWI have shown a close and direct relation to concurrent signals recorded by sEMG. MR-related magnetic fields had no significant effects on findings in sEMG. Hence, appearance of SMAMs in DWI should not be considered as imaging artifact or as effects originating from the special conditions of MR examinations. Spatial and temporal distributions of SMAMs indicate characteristics of spontaneous (microscopic) mechanical muscular action at rest. Therefore, DWI techniques should be considered as non-invasive tools for studying physiology and pathophysiology of spontaneous activities in resting muscle. Magn Reson Med 79:2784-2794, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Quick water-selective excitation of fast relaxing tissues with 3D UTE sequences

PURPOSE

The aim of this study was to implement a time effective 1-1 double pulse water-selective excitation (WE) into a three-dimensional ultrashort echo time (UTE) sequence (WE-UTE) for visualization of short-T2 tissues with positive contrast and sufficient suppression of surrounding fat.

METHODS

First, an analytical description of magnetization components in the steady state applying WE-UTE was derived and results were compared with numerical simulations based on Bloch's equations. Parameters were optimized for best positive contrast between short-T2 tissues and fat under consideration of variable relaxation properties over a broad range. Maximal signal yield and signal efficiency of on-resonant protons were compared with UTE sequences with and without off-resonance fat saturation (FatSat). WE-UTE was exemplarily applied for in-vivo musculoskeletal imaging on a 3T whole-body MR unit.

RESULTS

Steady state magnetization of WE-UTE could be described analytically and showed excellent accordance with numerical simulations. Even for tissues with T2 = 1 ms WE-UTE resulted in 79% of maximal signal yield of UTE without FatSat and was more efficient regarding signal yield if compared with UTE with FatSat. Using WE-UTE in-vivo tendons and ligaments could be well delineated with positive contrast to surrounding fat.

CONCLUSION

WE-UTE provides a quick method for visualizing short-T2 tissues with positive contrast.

Addressing spontaneous signal voids in repetitive single-shot DWI of musculature: spatial and temporal patterns in the calves of healthy volunteers and consideration of unintended muscle activities as underlying mechanism

Single-shot diffusion-weighted MRI sensitive to different types of incoherent motion inside tissue shows sporadic signal voids with a considerable size (>1 cm) in calf musculature at rest. Spatial and temporal patterns of these signal voids and their dependence on measurement conditions were tested systematically in order to obtain more insight into the underlying mechanism. Lower leg muscles of 10 healthy subjects were examined by recording series of 1000 echo-planar single-shot scans with repetition time 500 ms and b-value 100 s/mm(2) . Effects of strength and orientation of motion sensitization gradients and of repetition times were analysed. Potential influences of arterial blood pulsations and positioning of the subject were studied. Comparison of calf muscle groups showed more frequent signal voids in gastrocnemius and soleus muscle compared with tibialis muscles. Large inter-individual variance in the total number of signal voids visible in a transverse slice of the lower leg was observed (minimum 40/1000 scans; maximum >550/1000 scans). Typical sizes of the affected muscular areas ranged from 1.5 to 2.5 cm in the transverse and from 1.5 to 7 cm in the head-feet direction. Signal voids occurred nearly independent of the cardiac phase and with similar frequencies for supine and prone positions. Resting calf muscles show spontaneous signal voids in single-shot DWI at low b-values with an irregular temporal and spatial pattern. Values of mean diffusivity, diffusion tensor parameters, and IVIM-derived perfusion are expected to be clearly distorted by such signal voids if no rejection of affected data is applied. Several potential causes for the signal voids are discussed. The most probable explanation for the phenomenon is seen in the occurrence of spontaneous incoherent mechanical activity in musculature based on weak muscle fibre contractions. If this is the case it opens up a new field for studies on the physiological role and regulation of these unintended muscle activities.

A fully automatic reference deconvolution strategy to increase the accuracy of in vivo lipid signal quantification

PURPOSE

Lipid signals measured by (1)H MR spectroscopy cannot be adequately quantified by common fitting routines like VARPRO or AMARES, if lipid spectra are distorted by irregular spatial and temporal inhomogeneities of the static magnetic field during readout. A fully automatic reference deconvolution algorithm is presented that eliminates these distortions before application of fitting routines.

METHODS

The measured signal of the dominant methyl resonance is isolated with aid of a spectral estimator (estimation of parameters via rotational invariance techniques) and used as reference signal for estimation of distortions. A Wiener filter is applied to deconvolve those distortions in the lipid spectrum. Performance of the algorithm is assessed for different bandwidths and shapes of distortions, using artificially distorted as well as measured data.

RESULTS

Application of the fully automatic reference deconvolution algorithm on simulated spectra yields a distinct increase in quantification accuracy. Deconvolved in vivo spectra of subcutaneous fat indicate reduced spectral overlap after application of the proposed strategy.

CONCLUSION

The proposed method is helpful for in vivo magnetic resonance spectroscopy of adipose tissue to correct for effects of field inhomogeneities within the voxel and for inevitable eddy current effects. Quantification accuracy is improved by eliminating distortions before application of fitting routines.

Utilizing echo-shifts in k-space for generation of positive contrast in areas with marked susceptibility alterations

A technique for generation of positive contrast near susceptibility alterations utilizing echo-shifts in k-space is introduced, based on altered Larmor-frequencies and resulting phase-shifts accumulating during the echo-time at the site of local magnetic field gradients. 3D gradient-echo raw-data is acquired and weighted with an inverse Hanning filter. The filter partly suppresses central raw-data points, while maintaining outer areas. Reconstruction of the filtered raw-data results in images where pixels with apparent magnetic field gradients are highlighted against homogeneous pixels. Further processing steps are introduced to remove remaining intensities in the homogeneous parts of the filtered image. Feasibility is shown by an agar phantom containing magnetically labeled cells, with concentrations of 25, 50, 100, and 250 cells/μL, and by images of the human head. The technique allows detection of echo-shifted pixels with automatic suppression of magnetically homogeneous parts while keeping post-processing time short. Fewer than four labeled cells per pixel were clearly displayed with positive contrast. Application to the human head shows bright veins and complete suppression of homogeneous regions. The presented technique has high potential for specific detection of low concentrations of labeled cells or susceptibility altered regions in vivo with positive contrast, whereas areas with low spin density are not highlighted.

[Relaxation times T1, T2, and T2* of apples, pears, citrus fruits, and potatoes with a comparison to human tissues]

The aim of the project was a systematic assessment of relaxation times of different fruits and vegetables and a comparison to values of human tissues. Results provide an improved basis for selection of plant phantoms for development of new MR techniques and sequences. Vessels filled with agar gel are mostly used for this purpose, preparation of which is effortful and time-consuming. In the presented study apples, (malus, 8 species), pears, (pyrus, 2 species), citrus fruits (citrus, 5 species) and uncooked potatoes (solanum tuberosum, 8 species) from the supermarket were examined which are easily available nearly all-the-year. T1, T2 and T2* relaxation times of these nature products were measured on a 1.5 Tesla MR system with adapted examination protocols and mono-exponential fitting, and compared to literature data of human parenchyma tissues, fatty tissue and body fluid (cerebrospinal fluid). Resulting values were as follows: apples: T1: 1486-1874 ms, T2: 163-281 ms, T2*: 2.3-3.2 ms; pears: T1: 1631-1969 ms, T2: 119-133 ms, T2* : 10.1-10.6 ms, citrus fruits (pulp) T1: 2055-2632 ms, T2: 497-998 ms, T2* : 151-182 ms; citrus fruits (skin) T1: 561-1669 ms, T2: 93-119 ms; potatoes: T1: 1011-1459 ms, T2: 166 - 210 ms, T2* : 20 - 30 ms. All T1-values of the examined objects (except for potatoes and skins of citrus fruits) were longer than T1 values of human tissues. Also T2 values (except for pears and skins of citrus fruits) of the fruits and the potatoes tended to be longer. T2* values of apples, pears and potatoes were shorter than in healthy human tissue. Results show relaxation values of many fruits to be not exactly fitting to human tissue, but with suitable selection of the fruits and optionally with an adaption of measurement parameters one can achieve suitable contrast and signal characteristics for some purposes.

Positive contrast imaging of iron oxide nanoparticles with susceptibility-weighted imaging

Superparamagnetic iron oxide particles can be utilized to label cells for immune cell and stem cell therapy. The labeled cells cause significant field distortions induced in their vicinity, which can be detected with magnetic resonance imaging (MRI). In conventional imaging, the signal voids arising from the field distortions lead to negative contrast, which is not desirable, as detection of the cells can be masked by native low signal tissue. In this work, a new method for visualizing magnetically labeled cells with positive contrast is proposed and described. The technique presented is based on the susceptibility-weighted imaging (SWI) post-processing algorithm. Phase images from gradient-echo sequences are evaluated pixel by pixel, and a mask is created with values ranging from 0 to 1, depending on the phase value of the pixel. The magnitude image is then multiplied by the mask. With an appropriate mask function, positive contrast in the vicinity of the labeled cells is created. The feasibility of this technique is proved using an agar phantom containing superparamagnetic iron oxide particles-labeled cells and an ex vivo bovine liver. The results show high potential for detecting even small labeled cell concentrations in structurally inhomogeneous tissue types.

Quantitative diffusion imaging of adipose tissue in the human lower leg at 1.5 T

Diffusion-weighted spin-echo echo-planar imaging was developed and applied for assessment of diffusion coefficients of adipose tissue in human lower leg on a 1.5-T whole-body MR scanner. Because of the higher molecular weight of triglycerides, apparent diffusion coefficients (ADCs) of adipose tissue are approximately two orders of magnitude smaller compared with water, leading to the necessity of using high b-values up to 50,000 sec/mm(2) and an echo time of 240 msec for sufficient diffusion-related signal attenuation. ADC maps of adipose tissue in the human lower leg were derived for diffusion encoding along orthogonal spatial directions in six healthy volunteers. Mean diffusion coefficients in the tibial bone marrow amounted to (1.81 ± 0.10) × 10(-5) mm(2) /sec (left-right), (1.96 ± 0.10) × 10(-5) mm(2) /sec (anterior-posterior), and (1.96 ± 0.20) × 10(-5) mm(2) /sec (head-feet), respectively. Pixel-wise calculated ADC values of subcutaneous adipose tissue showed a distinctly higher variation with the smallest ADC values similar to those measured for tibial bone marrow. Some subcutaneous adipose tissue regions showed increased signal attenuation at higher b-values resulting in ADC coefficients up to 4.2 × 10(-5) mm(2) /sec. It must be noted that diffusion measurements with extremely high b-values in vivo are extremely sensitive to incoherent motion effects in tissue. Nonetheless, it could be shown that in vivo diffusion imaging of adipose tissue in human lower leg is possible at 1.5 T in acceptable measurement time of a few minutes. Potential future applications of fat diffusion imaging are seen in temperature measurements in adipose tissue, detection of free fatty acids in white or brown adipose tissue in case of high lipolysis, differentiation of macro- and microvesicular steatosis, or assessment of the mobility of intramyocellular lipids.

Effects of in-pulse transverse relaxation in 3D ultrashort echo time sequences: analytical derivation, comparison to numerical simulation and experimental application at 3T

The introduction of ultrashort-echo-time-(UTE)-sequences to clinical whole-body MR scanners has opened up the field of MR characterization of materials or tissues with extremely fast signal decay. If the transverse relaxation time is in the range of the RF-pulse duration, approximation of the RF-pulse by an instantaneous rotation applied at the middle of the RF-pulse and immediately followed by free relaxation will lead to a distinctly underestimated echo signal. Thus, the regular Ernst equation is not adequate to correctly describe steady state signal under those conditions. The paper presents an analytically derived modified Ernst equation, which correctly describes in-pulse relaxation of transverse magnetization under typical conditions: The equation is valid for rectangular excitation pulses, usually applied in 3D UTE sequences. Longitudinal relaxation time of the specimen must be clearly longer than RF-pulse duration, which is fulfilled for tendons and bony structures as well as many solid materials. Under these conditions, the proposed modified Ernst equation enables adequate and relatively simple calculation of the magnetization of materials or tissues. Analytically derived data are compared to numerical results obtained by using an established Runge-Kutta-algorithm based on the Bloch equations. Validity of the new approach was also tested by systematical measurements of a solid polymeric material on a 3T whole-body MR scanner. Thus, the presented modified Ernst equation provides a suitable basis for T1 measurements, even in tissues with T2 values as short as the RF-pulse duration: independent of RF-pulse duration, the 'variable flip angle method' led to consistent results of longitudinal relaxation time T1, if the T2 relaxation time of the material of interest is known as well.

Diffusion tensor imaging of the human calf muscle: distinct changes in fractional anisotropy and mean diffusion due to passive muscle shortening and stretching

The influence of passive shortening and stretching of the calf muscles on diffusion characteristics was investigated. The diffusion tensor was measured in transverse slices through the lower leg of eight healthy volunteers (29 +/- 7 years) on a 3 T whole-body MR unit in three different positions of the foot (40 degrees plantarflexion, neutral ankle position (0 degrees ), and -10 degrees dorsiflexion in the ankle). Maps of the mean diffusivity, the three eigenvalues of the tensor and fractional anisotropy (FA) were calculated. Results revealed a distinct dependence of the mean diffusivity and FA on the foot position and the related shortening and stretching of the muscle groups. The tibialis anterior muscle showed a significant increase of 19% in FA with increasing dorsiflexion, while the FA of the antagonists significantly decreased ( approximately 20%). Regarding the mean diffusivity of the diffusion tensor, the muscle groups showed an opposed response to muscle elongation and shortening. Regarding the eigenvalues of the diffusion tensor, lambda(2) and lambda(3) showed significant changes in relation to muscle length. In contrast, no change in lambda(1) could be found. This work reveals significant changes in diffusional characteristics induced by passive muscle shortening and stretching.

Influence of steady background gradients on the accuracy of molecular diffusion anisotropy measurements

Spatial susceptibility variations of body components lead to local gradients of the static magnetic field. Effects of such background gradients on fractional diffusion anisotropy (FA) measurements on whole-body magnetic resonance units operating at 1.5, 3.0 and 7.0 T were analyzed theoretically and experimentally. Analytical expressions were derived for the cases of diffusion occurring in isotropic media and in tissues with cylindrical symmetry (e.g., white matter tracts or skeletal musculature). Typical magnitudes of background gradient strengths were estimated from in vivo and in vitro measurements with B0 field mapping sequences. Additionally, numerical simulations of magnetic field distributions and resulting field gradients were performed considering tissue-air interfaces in simplified geometrical arrangements. For media with isotropic diffusion, both measurements and analytical calculations showed increasing FA inaccuracy with stronger coupling between diffusion-encoding and background gradients. For cylindrical symmetry, FA values were estimated for a standard diffusion tensor imaging protocol in a realistic scenario. At 1 mm distance from a water-air interface, susceptibility-related background gradients amount to approximately 9 mT/m at 7 T and lead to a relative error of the measured FA of up to 48%. The error in the anisotropy assessment rises considerably with increasing field strength and must be taken into account especially for experimental and clinical studies on modern high-field systems.

Heating of metallic implants and instruments induced by gradient switching in a 1.5-Tesla whole-body unit

PURPOSE

To examine gradient switching-induced heating of metallic parts.

MATERIALS AND METHODS

Copper and titanium frames and sheets ( approximately 50 x 50 mm(2), 1.5 mm thick, frame width = 3 mm) surrounded by air were positioned in the scanner perpendicular to the static field horizontally 20 cm off-center. During the execution of a sequence (three-dimensional [3D] true fast imaging with steady precession [True-FISP], TR = 6.4 msec) exploiting the gradient capabilities (maximum gradient = 40 mT/m, maximum slew rate = 200 T/m/second), heating was measured with an infrared camera. Radio frequency (RF) amplitude was set to zero volts. Heating of a copper frame with a narrowing to 1 mm over 20 mm at one side was examined in air and in addition surrounded by several liters of gelled saline using fiber-optic thermography. Further heating studies were performed using an artificial hip made of titanium, and an aluminum replica of the hip prosthesis with the same geometry.

RESULTS

For the copper specimens, considerable heating (>10 degrees C) in air and in gelled saline (>1.2 degrees C) could be observed. Heating of the titanium specimens was markedly less ( approximately 1 degrees C in air). For the titanium artificial hip no heating could be detected, while the rise in temperature for the aluminum replica was approximately 2.2 degrees C.

CONCLUSION

Heating of more than 10 degrees C solely due to gradient switching without any RF irradiation was demonstrated in isolated copper wire frames. Under specific conditions (high gradient duty cycle, metallic loop of sufficient inductance and low resistance, power matching) gradient switching-induced heating of conductive specimens must be considered.

Diffusion Tensor Imaging of the Spinal Cord at 1.5 and 3.0 Tesla

PURPOSE

The feasibility of highly resolved diffusion tensor imaging (DTI) of the human cervical spinal cord was tested on a clinical MR unit operating at 3.0 Tesla. DTI parametrical maps and signal-to-noise ratios (SNRs) were compared to results recorded at 1.5 Tesla.

MATERIALS AND METHODS

Eight healthy volunteers and one patient participated in the study. A transverse oriented single-shot ECG-triggered echo-planar imaging (EPI) sequence with double spin-echo diffusion preparation was applied for highly resolved DTI of the spinal cord. The signal yield, fractional anisotropy (FA), and mean diffusivity (MD) were compared for both field strengths. The clinical applicability of the protocol was also tested in one patient with amyotrophic lateral sclerosis (ALS) at 3.0 T.

RESULTS

A mean increase in SNR of 95.7 +/- 4.6 % was found at 3.0 Tesla compared to 1.5 Tesla. Improved quality of the DTI parametrical maps was observed at higher field strength (p < 0.02). Comparable FA and MD (reported in units of 10 (-3) mm (2)/s) values were computed in the dorsal white matter at both field strengths (1.5 T: FA = 0.75 +/- 0.08, MD = 0.84 +/- 0.12, 3.0 T: FA = 0.74 +/- 0.04, MD = 0.93 +/- 0.14). The DTI images exhibited diagnostic image quality in the patient. At the site of the diseased corticospinal tract, a decrease of 46.0 +/- 3.8 % in FA (0.40 +/- 0.03) and an increase of 50.3 +/- 5.6 % in MD (1.40 +/- 0.05) were found in the ALS patient.

CONCLUSION

The 3.0 Tesla field strength provides higher image quality in DTI of the spinal cord compared to 1.5 T. The proposed DTI protocol seems adequate for the assessment of spinal cord diseases.

Echoplanar diffusion tensor imaging of the lower leg musculature using eddy current nulled stimulated echo preparation

A sequence for echoplanar diffusion tensor imaging of musculature was developed using a stimulated echo preparation. The strategy was optimized in order to obtain reliable diffusion tensor data in a short measuring time. Image distortion problems due to eddy currents arising from long-lasting diffusion sensitizing gradients could be overcome by insertion of additional gradient pulses in the TM interval of the stimulated echo preparation. In contrast to former approaches with similar intention, the proposed strategy does not influence the stimulated echo signal itself and does not lead to prolonged echo time as in the case of spin echo methods. Phantom measurements were performed to compare eddy current induced distortion effects in diffusion weighted images. The diffusion tensor in the musculature of the lower leg was investigated in four healthy subjects and maps of the trace and the three eigenvalues of the diffusion tensor, fractional anisotropy maps, and angle maps were calculated.

Effects on MRI due to altered rf polarization near conductive implants or instruments

In magnetic resonance imaging near metal parts variations in radio frequency (rf)-amplitude and of receive sensitivity must be considered. For loop structures, e.g., vascular stents, B1 produces rf eddy currents in accordance to Faraday's law; the B1-related electrical rf field E1 injects directly to elongated structures (e.g., wires). Locally, the rf magnetic field Bl,ind (induced B1) is superimposed onto the rf field from the transmitter coil, which near the metal can dominate spin excitation. Geometry and arrangement of the parts determine the polarization of B(1,ind). Components parallel to B0 are of special interest. A copper sheet (100 mm x 15 mm, 3 mm thick) and a 27 cm long copper wire were examined in a water phantom using the spin-echo (SE) technique. In addition to rf-amplitude amplification, rf-phase shift due to z components of B(1,ind) could be detected near the metallic objects. Periodic rf-amplitude instabilities had an amplified effect for phase-shifted regions. Phase-encoding artifacts occurred as distinct ghosts (TR=200 ms) or band-like smearing (TR=201 ms) from affected spin ensembles. SE phase imaging can potentially be used in interventional magnetic resonance imaging for background-free localization of metallic markers.

rf enhancement and shielding in MRI caused by conductive implants: dependence on electrical parameters for a tube model

Radio frequency (rf) eddy-currents induced in implants made of conductive material might cause significant image artifacts in magnetic resonance imaging (MRI) such as shielding of the lumen of vascular stents. rf alteration near metal parts was assessed theoretically in the approximation of alternating current electrodynamics: The implant was modeled as tube with diameter d(o), resistance R, and reactance Y, constituting the secondary winding of a transformer. The transmitter coil of the scanner acted as primary winding and generated the linearly polarized rf field B1,app. Tube axis was assumed parallel to B1,app. The results of the calculations were as follows: Ninety percent of the applied rf-field amplitude is reached in the lumen at a ratio chi=R/Y approximately 2. A rapid drop occurs with the reduction of chi, whereas a further increase of chi causes only a small effect. With chi approximately 1/d(o)(Y approximately d2o,R approximately d(o)), conditions for rf alteration clearly depend on the diameter of the tube. Inside tubes with smaller diameter, rf shielding is less pronounced. rf alteration increases in good approximation with the square root of the strength of the static field B0. The following experiments were carried out: Tubes of similar diameter (d(o) approximately 8 mm) made of material of different conductivity (Cu, Nitinol, carbon fiber reinforced plastic with three different fiber structures) were examined at B0=0.2 and 1.5 T in water phantoms. Tube axis was aligned perpendicular to B0 and spin-echo technique was applied. Local rf enhancement near the outer surface of the metal tubes was detected applying manual reduction of the transmitter amplitude. Shielding inside a carbon fiber tube with d(o) approximately 8 mm and inside a smaller tube with d(o)=3.3 mm was compared. Both tubes showed the same wall structure and thickness (d(w)=0.4 mm). All measurements confirmed the theoretical results. Consequences for the construction of vascular stents are discussed, as well as problems with image artifacts due to rf enhancement near solid conductive implants.

Metal artifacts caused by gradient switching

In metal parts, e.g., implants or instruments, eddy currents can be induced from gradient switching if positioned off-center inside the MR scanner. For the first time, a systematic analysis of related artifacts was performed. Current strength increases in conjunction with increasing size of the part, increasing electrical conductivity, distance from isocenter, and increasing gradient strengths. A xy-plane oriented copper ring (d(o) = 20 mm, d(i) = 15 mm, 2 mm thick) was examined at isocenter and at x = 15 cm, y = z = 0. Comparisons of xy-, xz-, and yz-slices, recorded for both possibilities to select encoding directions, revealed effects from ramp-down of the slice-selection and ramp-up of the read-out gradient. Near the metal part, temporary inhomogeneities were superimposed to the static field and spin-dephasing signal loss resulted, despite using spin-echo technique. Artifacts depended on excitation and read-out bandwidth. For an equivalent titanium ring, conductivity related effects could not be ascertained but distinct susceptibility effects occurred. MR compatibility of implants/instruments therefore requires both low susceptibility and low conductivity.

Compensation of magnetic field distortions from paramagnetic instruments by added diamagnetic material: Measurements and numerical simulations

In minimally invasive procedures guided by magnetic resonance (MR) imaging instruments usually are made of titanium or titanium alloys (e.g., nitinol), because other more MR-compatible materials often cannot provide sufficient mechanical properties. Artifacts depending on susceptibility arise in MR images due to incorrect spatial encoding and intravoxel dephasing and thereby hamper the surgeon's view onto the region of interest. To overcome the artifact problem, compensation of the paramagnetic properties by diamagnetic coating or filling of the instruments has been proposed in the literature. We used a numerical modeling procedure to estimate the effect of compensation. Modeling of the perturbation of the static magnetic field close to the instruments reflects the underlying problem and is much faster and cost efficient than manufacturing prototypes and measuring artifact behavior of these prototypes in the MR scanner. A numerical model based on the decomposition of the susceptibility distribution in elementary dipoles was developed by us. The program code was written object oriented to allow for both maximum computational speed and minimum random access memory. We used System International units throughout the modeling for the magnetic field, allowing absolute quantification of the magnetic field disturbance. The field outside a simulated needlelike instrument, modeled by a paramagnetic cylinder (out of titan, chi =181.1) of length 8.0 mm and of diameter 1.0 mm, coated with a diamagnetic layer (out of bismuth, chi=-165.0) of thickness 0, 0.1, 0.2, 0.3, and 0.4 mm, was found to be best compensated if the cross-sectional area of the cylinder, multiplied by the absolute susceptibility value of the cylinder material, is equal to the cross-sectional area of the coating, multiplied by the absolute susceptibility value of the coating material. At the extremity of the coated cylinder an uncompensated field distortion was found to remain. We studied various tip shapes and geometries using our computational model: Suitable diamagnetic coating or filling of paramagnetic instruments clearly reduced tip artifacts and diminished the dependency of artifact size on orientation of the instrument with respect to B0 in the numerical studies. We verified the results of the simulations by measuring coated and uncoated titanium wires in a 1.5 T MR scanner.

Sodium 3-D MRI of the human torso using a volume coil

Sodium MR imaging is considered to provide clinically important information about the human body that is not achievable by hydrogen-based approaches. However, due to the low natural abundance in biological tissues, sodium signals usually lead to low spatial resolution, low SNR, and long acquisition times compared to conventional 1H imaging, even using well-adapted surface coils. For our study, a volume coil was designed with nearly homogeneous excitation/receive characteristics and a suitable geometry fitting the human torso. A sufficient penetration throughout the entire thorax, abdomen, or pelvis is provided allowing for sodium imaging of the kidneys, the liver with gall bladder, or the myocardium. All measurements were performed on a 1.5 T whole body scanner using a spoiled 3-D gradient echo sequence. Imaging parameters TE, TR, and readout bandwidth were optimized for sensitive recording of the sodium component with slow transverse relaxation. Nonselective RF excitation pulses with a duration of 2.5 ms and rectangular shape were applied to avoid SAR problems. Narrow receiver bandwidth and excitation near the Ernst angle provided clinically practicable examinations with measuring times of less than 15 min at a spatial resolution of 8 x 8 x 8 mm3. Under these conditions, SNR of 11 for the kidneys and vertebral disks, 9 for the spinal canal, and 6 for the liver was achieved. A special 3-D spin echo sequence was used to determine T2, times which resulted to 15.3 +/- 1.1 ms for liver, 27.7 +/- 7.2 ms for kidneys, and 24.0 +/- 4.7 ms for the content of the spinal canal.

Two-Dimensional Parallel Acquisition Technique in 3D MR Colonography

PURPOSE

In-vitro and in-vivo examinations to investigate the influence of one (1D)- and two-dimensional (2D) parallel acquisition techniques (PAT) on contrast-to-noise ratio (CNR) and image quality for bright-lumen 3D MR colonography.

MATERIALS AND METHODS

In-vitro measurements were performed on a 1.5 T whole-body MR scanner (MAGNETOM AVANTO(R), Siemens AG, Erlangen) with a standard spoiled 3D gradient-echo (3D GRE) sequence and a volume interpolated 3D GRE (VIBE) sequence using a home-built colon phantom. The relative CNR was determined and image quality evaluated for different acceleration factors (PAT factors). A bright-lumen 3D MR colonography with PAT factors 2, 3, 4 and 6 was performed on a 39-year-old volunteer. The 3D data sets were compared with regard to image quality.

RESULTS

The mean relative CNR values for the in-vitro measurements were 1, 0.81, 0.73, 0.52 and 0.4 (3D GRE) and 1, 0.8, 0.65, 0.45 and 0.3 (VIBE) for the PAT factors 1, 2, 3, 4 and 6, respectively. Residual aliasing artifacts not affecting image quality in a distinct manner were visible for 2D PAT, especially for the VIBE sequence. Increasing the PAT factor up to 6 and decreasing the acquisition time down to 10 seconds for the bright-lumen 3D MR colonography could achieve adequate image quality with significantly reduced image artifacts caused by peristalsis and pulsations.

DISCUSSION

Even for high PAT factors up to 6, 2D PAT only leads to a moderate CNR loss. For a 3D MR colonography, distinct shorter acquisition times can be achieved with identical resolution.

Numerical modeling of needle tip artifacts in MR gradient echo imaging

Exact determination of needle tip position is obsolete for interventional procedures under control of magnetic resonance imaging (MRI). Exact needle tip navigation is complicated by the paramagnetism of microsurgical instruments: Local magnetic field inhomogeneities are induced resulting in position encoding artifacts and in signal voids in the surrounding of instruments and especially near their tips. The artifacts generated by the susceptibility of the material are not only dependent on the material properties themselves and on the applied MRI sequences and parameters, but also on the geometric shape of the instruments and on the orientation to the static magnetic field in the MR unit. A numerical model based on superposition of induced elementary dipole fields was developed for studying the field distortions near paramagnetic needle tips. The model was validated by comparison with experimental data using field mapping MRI techniques. Comparison between experimental data and numerical simulations revealed good correspondence for the induced field inhomogeneities. Further systematic numerical studies of the field distribution were performed for variable types of concentric and asymmetric tip shapes, for different ratios between tip length and needle diameter, and for different orientations of the needle axis in the external static magnetic field. Based on the computed local inhomogeneities of the magnetic field in the surroundings of the needle tips, signal voids in usual gradient echo images were simulated for a prediction of the artifacts. The practically relevant spatial relation between those artifacts and the hidden tip of the needle was calculated for the different tip shapes and orientations in the external field. As needle tip determination is crucial in interventional procedures, e.g., in taking biopsies, the present model can help to instruct the physician prior to surgical interventions in better estimating the needle tip position for different orientations and needle tip shapes as they appear in interventional procedures. As manufacturing prototypes with subsequent measurements of artifacts in MRI are a costly procedure the presented model may also help to optimize shapes of needle tips and of other parts of MR-compatible instruments and implants with low expense prior to production if some shape parameters can be chosen freely.

Separation of intra- and extramyocellular lipid signals in proton MR spectra by determination of their magnetic field distribution

In skeletal musculature intramyocellular (IMCL) and extramyocellular lipids (EMCL) are stored in compartments of different geometry and experience different magnetic field strengths due to geometrical susceptibility effects. The effect is strong enough to---at least partly---separate IMCL and EMCL contributions in (1)H MR spectroscopy, despite IMCL and EMCL consisting of the same substances. The assessment of intramyocellular lipid stores in skeletal musculature by (1)H MR spectroscopy plays an important role for studying physiological and pathological aspects of lipid metabolism. Therefore, a method using mathematical tools of Fourier analysis is developed to obtain the magnetic field distribution (MFD) from the measured spectra by deconvolution. A reference lipid spectrum is required which was recorded in tibial yellow bone marrow. It is shown that the separation of IMCL contributions can be performed more precisely---compared to other methods---based on the MFD. Examples of deconvolution in model systems elucidate the principle. Applications of the proposed approach on in vivo examinations in m. soleus and m. tibialis anterior are presented. Fitting the IMCL part of the MFD by a Gaussian lineshape with a linewidth kept fixed with respect to the linewidth of creatine and with the assumption of a smooth but not necessarily symmetrical shape for the EMCL part, the only free fit parameter, the amplitude of the IMCL part, is definite and subtraction leads to the EMCL part in the MFD. This procedure is especially justified for the soleus muscle showing a severely asymmetrical distribution which might lead to a marked overestimation of IMCL using common line fitting procedures.