MRI (magneticresonanceimaging) is an indispensable tool in the diagnosis of centrals nervous system (CNS) disorders such as spinal cord injury and multiple sclerosis (MS). In contrast, diagnosis of peripheral nerve injuries largely depends on clinical and electrophysiological parameters. Thus, currently MRI is not regularly used which in part is due to small nerve calibers and isointensity with surrounding tissue such as muscles. In this study we performed translational MRI research in mice to establish a novel MRI protocol visualizing intact and injured peripheral nerves in a non-invasive manner without contrast agents. With this protocol we were able to image even very small nerves and nerve branches such as the mouse facial nerve (diameter 100–300 µm) at highest spatial resolution. Analysis was performed in the same animal in a longitudinal study spanning 3 weeks after injury. Nerve injury caused hyperintense signal in T 2 -weighted images and an increase in nerve size of the proximal and distal nerve stumps were observed. Further hyperintense signal was observed in a bulb-like structure in the lesion site, which correlated histologically with the production of fibrotic tissue and immune cell infiltration. The longitudinal MR representation of the facial nerve lesions correlated well with physiological recovery of nerve function by quantifying whisker movement. In summary, we provide a novel protocol in rodents allowing for non-invasive, non-contrast agent enhanced, high-resolution MR imaging of small peripheral nerves longitudinally over several weeks. This protocol might further help to establish MRI as an important diagnostic and post-surgery follow-up tool to monitor peripheral nerve injuries in humans.
Thus, the choice of using onsite or shared MRI equip-
ment is a variant of the classic “make or buy” decision in
organizational economics, which is mainly influenced by scope economies and transaction economies [23–29]. The make-or-buy decision has been studied extensively in the context of transaction cost economics, which posits that the boundaries of organizations are in large part a function of the nature of the business transacted, where relatively complex transactions are more effi- ciently organized in settings that feature stronger admin- istrative controls, such as ownership [23, 27, 29]. In the market for medical care, consumer transaction costs are the costs incurred to the consumer to complete a trans- action, including the time necessary to implement in- formed choice, such as evaluating, choosing and locating a care provider, as well as the time spent directly obtain- ing the services . Consumer transaction costs are ex- pected to be lower in the case of onsite MRI availability because patients may be able to economize on identify- ing, vetting, locating and traveling to a provider . In addition, there are several potential convenience-related benefits associated with onsite availability, including
For Subproject 3 (Publications II and VII), patients were included in a prospective imaging study (7.0 Tesla Ultra-High Field Project, “7UP-Study”, WHO International Clinical Trials Registry No. DRKS00003193, http://apps.who.int/trialsearch/Trial.aspx?TrialID=DRKS00003193). Inclusion criteria were: (1) subacute/chronic ischemia or transitory ischemic attack (TIA), (2) age 18–80 years, (3) ability to give informed consent and (4) legal competence. Exclusion criteria were: (1) cardiac pacemakers or any other electronic implants, (2) any metallic implant, (3) pregnancy or breast feeding period, (4) claustrophobia, (5) chronic or episodic vertigo, (6) retinal diseases and (7) dental bridges and more than two metallic dental crowns in a row. Neurological status was assessed using the National Institute of Health Stroke Scale (NIHSS) at time of admission for index stroke and before MR imaging. Imaging was performed first at 3.0 T, immediately followed by 7.0 T. A neurologist specialized in stroke supervised the patients during MRI. In total, 18 patients were eligible for Publication II and 6 patients for Publication VII, which was performed on the subset of Moya-Moya-patients (n=6).
The novel technique for phase-based quantitative conductivity mapping finds it application in brain tumour management, as demonstrated in Chapter 5. Furthermore, a novel method to estimate and quantify tissue sodium concentration in the brain in vivo which relies on MR-based electrical properties mapping (MR-EPT) is proposed in Chapter 6. To validate this approach, conductivity maps and sodium concentration maps of the brain (the latter one being based on sodium MRI) were estimated independently using state-of-the-art approaches in a group of 8 healthy adults. The so-called pseudo tissue sodium concentration (pTSC) maps derived from the conductivity map using the proposed model was compared with the “true” sodium concentration map (TSC) and the agreement between the two was studied. A statistically significant Pearson correlation (p<0.001; r=0.6) was observed between the two modalities while Bland-Altman analysis revealed a discrepancy of ~4mMol/L on average over the whole brain. This study suggests that pseudo tissue sodium concentration mapping could become a valuable alternative method to infer quantitative sodium concentration maps in the brain, and directions of improvements are discussed to lead to progress in this field, namely the possibility to include tissue water content to refine the model. A perspective technological application of this work is the possibility to estimate sodium concentration maps at sites where the necessary hardware for sodium MRI is not available. This work is revised and accepted in the journal “MagneticResonance in Medicine”. Consequently, Chapter 6 is an adaptation of the original article with modifications and unpublished results.
In the previous sections we described the response of spins to an ex- ternal magnetic field and the resulting NMR signal. The described raw signal — the free induction decay — does not yield any infor- mation about the location of the spins. For the generation of an image one needs information on spin location. This information can be gained by using time–varying gradient fields. The gradient fields used in imaging are generally spatially linear varying magnetic fields which are superimposed on the static magnetic field. In an MRI scanner these fields are generated by electrical gradient coils driven by powerful amplifiers which enable a rapid switching of the gradi- ents. Compared to the magnetic field which is on the order of 1.5-3 Tesla for clinical scanners and up to 9.4 Tesla for state–of–the–art research instruments the magnetic field gradients have relatively low maximum field strengths usually on the order of 80 mT with gradient strengths of 40-80 mT m for whole body scanners.
One observer (G.D.), with 3 years’ experience of fetal MRI, evaluated all subjects in both groups. The following parameters were assessed quantitatively using the open-source software application ITK-SNAP (www .itksnap.org), version 3.6 17 (Figure 1): number of differ- entiable vermian lobules, BV angle, midsagittal vermian area, area of brainstem structures (mesencephalon, pons and medulla oblongata), presence of vermian tail sign, presence of external mechanical pressure on the vermis, differentiability of the declive, folium and tuber (DFT) and differentiability of the uvula from the nodulus. The
On the other hand, MRI offers the possibility of obtaining relaxation-time weighted images as well as relaxation time maps that allow conclusions to be drawn about local water mobility. In the context of NMR, water mobility means the molecular mobil- ity based on rotational and translational diffusion. These motions create fluctuating local magnetic fields responsible for the relax- ation processes. In porous media, the local mobility of water can be restricted by adsorption at solid–liquid interfaces, confinement in narrow structures, incorporation in gels, or exchange among bulk, surface, and polymer phases (Belton, 1997; Brownstein and Tarr, 1977, 1979; Kimmich, 2012). Consequently, short relaxation times occur for water molecules residing in small pores (Kleinberg et al., 1994), in films at low water contents (Haber-Pohlmeier et al., 2014), or if water is incorporated in gels (Bertasa et al., 2018; Brax et al., 2017). When mucilage from chia seed ( Salvia hispanica L. ) is mixed with soil, the situation becomes more complex. Brax et al. (2018) proposed a so-called gel effect characterized by accelerated bulk and surface relaxation even at constant pore size and water content. In a recent work, van Veelen et al. (2018) reported about relaxation time mapping at high field using the transverse relax- ation time instead of the longitudinal relaxation.
Since the mask should be related to the phase robustness or quality, it makes sense to employ a quality measure. A beneficial choice is e.g. the LC measure (Section 4.7.1, Witoszynskyj et al.
[ 2009 ]). The measure reflects how strongly the phase of voxels differs in a local neighbourhood. It is maximal, if all voxels have the same phase. For an MRI measurement the phase within regions with low signal is typically randomised, thus the LC will be extremely low. In regions with sufficient signal, the behaviour of the LC still depends on phase behaviour: For phase varying smoothly with regard to the local neighbourhood dimensioning the LC will be close to 1, whilst for neighbourhoods with rapid phase changes, the LC decreases towards 0. If a small neighbourhood, such as 3 × 3 × 3 voxels, is chosen, these characteristics become quite useful. One wants to exclude noise, but also local variations that are likely to create phase sampling problems. Thus, exclusion of voxels with a low LC even inside a measured object, can be advantageous. Due to statistics it is possible, that in a few neighbourhoods residing in noisy areas the LC is accidentally high. E.g. random coherences of a major fraction of these voxels can have this effect. In order to avoid accidental inclusions of voxels in the signal void, a gentle smoothing filter, such as a Gaussian can be applied. Finally the smoothed LC is thresholded in order to obtain a mask for the region of valid phase. A visualisation of the steps involved in this method is shown in Fig. 4.21. Additionally, it is advisable to identify the biggest continuous region in the resulting mask. This avoids the inclusion of isolated and falsely assigned regions.
As a gold-standard control, each bone sample was scanned with µCT (SkyScan 1272; Bruker, Kontich, Belgium) with 80 kV, 125 mA, and a voxel size of 16 µm. For MRI scanning, each sample was inserted into a plastic conical container filled with ultrasound gel. T1-weighted turbo spin echo sequence imaging was performed using a whole-body 3T magnetic-resonance system (Philips, Healthcare System) with 8-channel SENSE-foot/Ankle coil, TR 25 ms, TE 3.5 ms, FOV 100 × 100 × 90, voxel size 0.22 × 0.22 × 0.50 mm, and a scan time of 11:18. Tissue with short T 2 relaxation times was presented as bright values.
All patients underwent MRI of the spine to assess the level and degree of the suspected fracture and acute bone-marrow edema. The fracture site and bone mar- row edema were defined by two radiologists with 10 and 7 years of experience in musculoskeletal imaging in consensus. Acute osteoporotic fracture was diagnosed ac- cording to MR-specific criteria and correlation with additionally available studies, including conventional imaging and computed tomography. MR-specific criteria in- dicating acute fracture were a hyperintense fluid-like signal within a visible fracture cleft on T2-weighted TSE and STIR-imaging as well as a hyperintense bone-marrow edema within the affected vertebra, which corresponded to a hypointense signal of the affected bone marrow compared to the physiological hyperintense bone-marrow fat signal in T1-weighted TSE-sequences. In order to differentiate the fracture types on the morphological images, established criteria describing the osteoporotic nature of a fracture were used, such as the fluid sign or ground- and endplate impression fracture pattern. Signs indicating a malignant fracture or infiltration included solid, enhancing soft-tissue components or clear signs of infiltration into the vertebral arch, prominent posterior vertebral bulging or paravertebral or intraspinal infiltration.
Utilizing the Akaike Information Criterion (AIC), the acquired data was shown to be best de- scribed by a two-compartment model consisting of a static and a flowing compartment, where the static compartment is hypothesized to be ascribed to liquid inside the spherical dilations. The flow- ing compartment was fit using the phase-distribution model. In accordance with the expectation that the relative volumes of multiple compartments should be independent of the applied flow rate, the determined signal fraction 𝑓 attributed to the flowing compartment stayed approximately con- stant over the course of the measurement series. The average signal fraction 𝑓 = 0.451 ± 0.023 of all flow rates from the measurement series A is appreciably close to the ratio of the volume inside the capillaries 𝑉 𝑐𝑎𝑝 to the total network volume 𝑉 𝑐𝑎𝑝 + 𝑉 𝑑𝑖𝑙 , estimated at 0.454 ± 0.002 using the data from optical microscopy. This further supports the notion that the static compartment consists of the liquid inside the spherical dilations. Having a flowing and static compartment of roughly the same size benefits the detailed analysis of flow effects using DW-MRI; however, with respect to in vivo measurements, a perfusion fraction of 𝑓 = 0.451 can be considered rather large. Values close to that have only been reported in the best-perfused organs such as the kidneys 104 or the liver 27 . In a closed system, the average particle speed due to flow must be linearly proportional to the applied rate of flow. The estimated average particle speed 〈𝑣〉 inside the capillary phantom fully met this expectation. The calculated coefficient of determination between the applied rate of flow and 〈𝑣〉 is 𝑅 2 = 0.99, indicating that practically all variation in 〈𝑣〉 is predictable from the applied
thickness measurements from MRI images are similar to a measurement made on the front feet of normal horses included in an MRI study of chronic laminitis (MURRAY et al. 2003). The MRI data are a possible reference for clinical and research MRI studies of equine feet. The primary difference between DR and MR measurements was better tissue contrast resolution noted on MR images, which may allow for more precise positioning of calipers at the tissue interfaces when making measurements. This was important, because it allowed us to distinguish the tissue components of the two soft tissue layers (outer opaque and inner more lucent) of the hoof wall, which can be delineated on DR.
The liquid phase mobility in the monoliths with 3.16 wt% IL and 3.20 wt% stabilizer (Fig. 7a) as well as 3.23 wt% IL and 3.28 wt% stabilizer (Fig. 7b) were examined by impregnating the monoliths horizontally by complete or partial (ca. 50% coverage) submersion in stock solution followed by drying. Notably, the positions of the monoliths were maintained during the impregnation and the MRI analysis, ensuring that the bottom of the images were facing towards the center of gravity. During both horizontal impregnation procedures, the mono- liths were found to readily soak up the stock solution due to good wettability of polar solvents (like DCM) on SiC 43 resulting in 6.4 and 6.5 wt% uptake, respectively. In line with this, images of both monoliths revealed complete impregnation of the cross- sections independent of gravity with only minor accumulation of species in the lower right part (Fig. 7a) and the upper rim (Fig. 7b), respectively. This rmly demonstrated that the orientation of the monolith during impregnation was not a key
A major technical challenge in coil array design is the mutual decoupling between individual coil elements. Conventional decoupling techniques use either geometrical overlap , with the drawback of reduced overall FOV and higher g-factors for parallel imaging due to the overlapping sensitivity profiles. Another decoupling strategy includes LC-networks between nonoverlapping coils [63,65], with the disadvantage of frequency and load-dependent decoupling efficiency. Some authors proposed strategies to decouple physically separated coils by magnetic flux sharing to combine the advantages of overlap and LC-network decoupling. Avdievich and Hetherington  used a pair of overlapping annex loops with opposite winding orientation connected in series with two neighboring surface coil elements. Constantinides and Angeli  placed closed copper loops proximal to the array, partially overlapping with the mutually interacting surface coils, and thereby eliminating the magnetic coupling. Low-impedance preamplifiers are widely used for inter-element decoupling in receiver arrays . In transmit arrays, the mutual coupling can be reduced with the current source RF amplifier method [61,62], although it is currently not available for most MRI systems.
of treatment with citalopram (n = 32, 1.5 T; 6-week measure- ment interval) ( Arnone et al., 2013 ). This study also found larger hippocampal volumes in patients who were remitted, although not in the same pattern as we found (aMDD < HC < rMDD in our study vs aMDD < rMDD < HC in Arnone et al.). A higher number of episodes in their remitted sample might explain this discrep- ancy, and this study was better matched for sex and age than in our sample. Hippocampal volume increases after paroxetine treatment were found earlier in posttraumatic stress disorder (n = 23, 1.5 T, 12 months ( Vermetten et al., 2003 ). While the first study used a voxel-based morphometry approach with masks, the latter applied manual delineation, which are methodologic- ally different to newer methods ( Cao et al., 2017 ; Maller et al., 2018 ). Absent volume increases in the total hippocampus as well as absent total hippocampal differences between depressed Figure 3. Exploratory analysis of hippocampal subfield differences before treatment in acute depressed patients according to remitter status after 12 weeks of treat- ment (n = 10/10). Larger subfield volumes were found in the presubiculum, right fissure, and right fimbria in nonremitting depressed patients (aMDDnon_rem) com- pared with remitting patients (aMDDrem) before treatment. No significant changes were obtained between MRI-1 and MRI-2 or at MRI-2. **P < .05 corrected with Tukey’s method, *P < .05, uncorrected.
From each patella a cylindrical cartilage-on-bone sample of 14 mm diameter was drilled perpendicular to the articular surface from the centre of the medial patellar facet. During drilling, the patella was continuously rinsed with cooled (5 °C) physiological saline. The desired MRI slice was visually selected and marked with an incision in the bony part of the sample which could be seen in MRI. The sample was placed in a cylinder of acrylic glass (plexiglas, 20 mm outer diameter) tightly fitting into the coil to prevent motion artifacts. The inner diameter of the cylinder was equal to the diameter of the sample and had a thread of 1 mm pitch matching with a cylindrical screw for loading experiments. After inserting the sample in the plexiglas cylinder, physiologic saline was added to prevent cartilage dehydration during MRI. The sample was positioned so that its articular surface was oriented perpendicular to the external magnetic field.
Material & Methods: LN in preoperative MRI of 60 patients with rectal cancer and primary surgery that received no neoadjuvant chemoradiation were measured and characterized in terms of border contour (smooth/unaltered, altered border contour in general, lobulated or spiculated/indistinct) and homogeneity of signal intensity (homogeneous or inhomogeneous). Using the size criterion for LN metastasis (cN+) LN of 7,2 mm or more in diameter were considered metastatic. A spiculated/indistinct border contour and inhomogeneous signal intensity were signs of lymph node metastasis (cN+) when applying morphological criteria. Accuracy of these signs for clinical LN staging were analyzed with conclusive postoperative histological lymph node examination and compared to each other.
Publikation 3: Tahir Durmus, Carsten Stephan, Maria Grigoryev, Gerd Diederichs, Musaab Saleh, Torsten Slowinski, Andreas Maxeiner, Anke Thomas, Thomas Fischer; Detection of prostate cancer by real-time MR/Ultrasound fusion guided biopsy: 3T MRI and state of the art sonography techniques; Rofo. 2013 May;185(5):428-33.: Dateninterpretation. Verfassen einzelner Passagen und kritisches Überarbeiten des Manuskripts. Finale Zustimmung der zu publizierenden Version des Artikels
Advanced medical imaging techniques require high performance segmentation algorithms. The main challenge in medical image segmentation tasks is to extract accurately the structures of interest. In the case of medical imaging the segmentation process can take place in 2D or in the 3D image domain. Usually 2D methods are applied on a single-image dataset and 3D methods are used for volume segmentation. Nevertheless, 2D algorithms can be extended to 3D medical volumes by being applied successively on the compounding 2D slices    . The last approach is in some cases more practical as it is easier to implement, it requires less memory, and has lower computation complexity. This section gives an overview of the 2D and 3D most common segmentation algorithms applied in medical imaging. An overview for the implementation of each method is given and the advantages and disadvantages are discussed. It has to be mentioned that often, depending on the task, several algorithms are combined together with a view to solve difficult segmentation tasks.
For each MRI examination, readers had to evaluate pres- ence or absence or foci. The BI-RADS definition of a focus was strictly followed [ 11 ], and lesions that could be further characterized by evaluating pre- and post-contrast T1w sequences, T2w sequences, or DWI were not considered, even when presenting a diameter equal or lower to 5 mm (i.e., cysts, small spiculated masses, and lymph nodes). In patients showing strong, punctate, background enhance- ment, a focus was described only when showing enhanc- ing characteristics clearly different from that of the remain- ing fibroglandular tissue. For all detected foci, readers had to state whether in the subsequent examinations, the lesion was disappearing, reducing in size, stable, increas- ing in size, or showed any change in morphology sug- gesting malignancy. Number of foci per patient was also considered.