3D Free-breathing multichannel absolute Mapping in the human body at 7T.

This thesis is about feature-based mobile robot navigation in unknown environments. The work is focusing on solving the problem of Simultaneous Localization and Mapping (SLAM) including feature extraction, estimation and complete map building solution. The robot platforms used in this work have the encoders for odometry and several laser scanners for exteroperceptive sensors. Simulations, real experiments and existing well known datasets are used in order to verify the performance of the developed algorithms. The algorithms developed in this thesis are features based meaning that instead of directly using raw sensory data geometric features are first extracted. This allows the filtering of noise from the sensors and the dynamics of the environment. More importantly, feature based approaches can reduce the amount of data needed to process, enabling the algorithms to scale in large environments. Another advantage of using features is the characteristic of better isolating informative patterns so that by exploiting specific knowledge of the environment, specialized algorithms can be easily adopted with different types of features. Several geometric features (points, lines, planes) are considered in this work. In this thesis, the relative mapping approach is adopted for solving the SLAM problem. This approach has many advantages. First, the mapping is decoupled from the estimation of robot pose so that the map estimation is independent from odometry errors. In other words, odometry errors do not affect and accumulate to the map estimate as in the absolute mapping approach. Second, the estimation does not involve the dynamics model of the robot which is identity in a relative mapping algorithm, meaning that only the observation model is considered in the estimation. A relative map has extra consistency problem compared to an absolute map. It is subject to the geometrical consistency problem which may generate an inconsistent absolute map after an transformation. However, having only observation model in the estimation gives an insight to separating nonlinearities from the iterative part of the estimator and thus minimize the effects of linearization errors. In this work, these subjects are investigated and reported. For some service robotic applications, especially for indoor office-like environments, we would like to have a fast, robust and lightweight approach which is capable to perform real-time SLAM on minimal hardware platforms. By exploiting some known knowledge about the structure of the environment, we claim to be able to achieve that goal. Particularly, the OrthoSLAM algorithm has been developed for building 2D and 3D maps. The algorithm adopts the Orthogonality assumption (or constraint) on the environment. First, the complexity of the SLAM algorithm is significantly reduced. Second, the linearization error from the robot heading uncertainty is removed and thus the map consistency is greatly improved. Implementations of the developed algorithms are tested on simulated data, on existing real world datasets and on robot platforms. Existing algorithms used in this work have been also reimplemented. The performance of the algorithms are analysed and compared on well chosen criteria. Particularly, the issues on computational complexity, linearization error, map consistency, map precision, realtime capability are investigated.

Quantitative Magnetic Resonance Imaging of the epicardial adipose tissue with free-running cardiac Dixon

Human heart imaging at ultra-high fields is highly challenging because of respiratory motion-induced artefacts and spatially heterogeneous profiles. This work demonstrates that respiration resolved 3D -maps can be used with a dedicated tailored and universal parallel transmission (pTx) pulse design to compensate respiration related changes in subjects performing shallow and deep breathing (SB/DB). Three-dimensional (3D) -maps of the thorax were acquired in 31 subjects under SB and in 15 subjects under SB and DB. Different universal and tailored non-selective pTx pulses were designed from non-respiration resolved (NRR) and respiration resolved (RR) reconstructions of the SB/DB -maps. The performance of all pulses was tested with RR-SB/DB -maps. Respiration-robust tailored and universal pulses were applied in vivo in 5 subjects at 7T in 3D gradient-echo free-breathing scans. All optimized pTx pulses performed well for SB. For DB, however, only the universal and the tailored respiration-robust pulses achieved homogeneous flip angles (FAs) in all subjects and across all respiration states, whereas the tailored respiration-specific pulses resulted in a higher FA variation. The respiration-robust universal pulse resulted in an average coefficient of variation in the FA maps of 12.6% compared to 8.2% achieved by tailored respiration-robust pulses. In vivo measurements at 7T demonstrate the benefits of using respiration-robust pulses for DB. Universal and tailored respiration-robust pTx pulses based on RR -maps are highly preferred to achieve 3D heart FA homogenization at 7T when subjects perform DB, whereas universal and tailored pulses based on NRR -maps are sufficient when subjects perform SB.

To design low peak and integrated power simultaneous multislice excitation radiofrequency pulses with transmit field inhomogeneity compensation in high field MRI. The "interleaved greedy and local optimization" algorithm for small-tip-angle spokes pulses is extended to design multiband (MB) spokes pulses that simultaneously excite multiple slices, with independent spokes weight optimization for each slice. The peak power of the pulses is controlled using a slice phase optimization technique. Simulations were performed at 7T to compare the peak power of optimized MB spokes pulses to unoptimized pulses, and to compare the proposed slice-independent spokes weight optimization to a joint approach. In vivo experiments were performed at 7T to validate the pulse's function and compare them to conventional MB pulses. Simulations showed that the peak power-minimized pulses had lower peak power than unregularized and integrated power-regularized pulses, and that the slice-independent spokes weight optimization consistently produced lower flip angle inhomogeneity and lower peak and integrated power pulses. In the brain imaging experiments, the MB spokes pulses showed significant improvement in excitation flip angle and subsequently signal homogeneity compared to conventional MB pulses. The proposed MB spokes pulses improve flip angle homogeneity in simultaneous multislice acquisitions at ultrahigh field, with minimal increase in integrated and peak radiofrequency power.

Highly undersampled three-dimensional (3D) saturation-recovery sequences are affected by k-space trajectory since the magnetization does not reach steady state during the acquisition and the slab excitation profile yields different flip angles in different slices. This study compares centric and reverse-centric 3D cardiac perfusion imaging. An undersampled (98 phase encodes) 3D ECG-gated saturation-recovery sequence that alternates centric and reverse-centric acquisitions each time frame was used to image phantoms and in vivo subjects. Flip angle variation across the slices was measured, and contrast with each trajectory was analyzed via Bloch simulation. Significant variations in flip angle were observed across slices, leading to larger signal variation across slices for the centric acquisition. In simulation, severe transient artifacts were observed when using the centric trajectory with higher flip angles, placing practical limits on the maximum flip angle used. The reverse-centric trajectory provided less contrast, but was more robust to flip angle variations. Both of the k-space trajectories can provide reasonable image quality. The centric trajectory can have higher CNR, but is more sensitive to flip angle variation. The reverse-centric trajectory is more robust to flip angle variation.

Motivation: Last year we demonstrated at 10.5 T the utility of motion-robust multi-echo GRE imaging for mesoscale T2*-weighted (T2*w) whole brain MRI with 0.5-mm isotropic resolution. Goal(s): To push the limit of mesoscale T2*w multi-echo whole brain MRI by combining motion-robust 3D EPI and parallel transmission (pTx). Approach: Whole brain three-echo EPI at 0.4-mm isotropic resolution was performed using a custom 16-channel transmit 80-channel receive RF array with a scan time of ~10 min. Tailored pTx pulses were designed for uniform water excitation. Results: Our new mesoscale image acquisition eliminated RF artifacts observed before, allowing intricate anatomical details to be visualized across the brain. Impact: Armed with rapid motion-robust 3D EPI and parallel transmission, the presented new multi-echo T2*-weighted imaging strategy will benefit many neuroscience applications, especially those aiming to investigate fine scale anatomy in humans at ultrahigh field.

In MRI-guided needle interventions, such as biopsies and brachytherapy, tissue deformation caused by needle movement may result in localization errors and thus hamper the outcome of the procedure. Monitoring the local tissue deformation provides the ability to compensate for it, e.g., by increasing the needle insertion depth. Fast MR scans are useful to track the needle, but cannot be used to quantify local tissue deformation, in case anatomical landmarks are missing. Artificial landmarks can be created by MR tagging. This method provides a spatial saturation pattern (tag) in the tissue. Deformation of this pattern reflects the tissue motion between tag creation and tag imaging. As the needle movement is nonperiodic, k-space cannot be acquired with a multishot approach, like is usually done for cardiac imaging. Hence, a single shot MR tagging sequence is needed, which entails tag creation, needle movement and tag acquisition. In this study, the feasibility of single shot MR tagging for MRI-guided needle interventions in phantom and volunteer experiments is shown. Four different experiments were performed on a 1.5 T MR scanner: the first to quantify translations, the second to quantify rotations, the third to mimic a needle intervention, and the fourth to investigate the tag persistence in a volunteer. The tag pattern is created by a 1331 composite pulse. A balanced steady state free precession sequence is used for imaging. To minimize undesired changes in contrast or sharpness of the tag pattern, we chose a relatively small flip angle and a short imaging time in all experiments. In the volunteer experiments, we modified the sequence to also be able to inspect the influence of the used k-space sampling profile and the flip angle on the temporal persistence of image contrast and tag pattern. In all scans, head or surface coils were used for signal reception. In all experiments, the tag pattern was clearly visible and could be used to quantify the local tissue deformation caused by (needle) movement. Strong correlations between the actual and measured (angular) phantom motions were obtained. In the needle intervention experiment, the tag lines were perfectly horizontal when there was no needle movement. With needle movement, local tissue displacements up to 5 mm were observed. Volunteer's anatomy could be discriminated, despite the tag pattern. The tag pattern in the prostate, for example, could still be read in all tagging images acquired 2 s after creating the tag pattern. With optimized scan parameters the tag persistence was even longer. The best image tag contrast was obtained using a large flip angle and the profile order low-high, although the image was slightly blurred. This study demonstrates that single shot MR tagging can be used to quantify tissue deformation caused by needle movement. The in-vivo tag persistence is sufficient to enable the application of the tagging sequence during MRI-guided needle interventions in patients.

To study in simulation and in theory the accuracy and precision of dynamic T1 measurements obtained using the previously published single-reference variable flip angle (SR-VFA) technique, with a focus on the effects of dynamic changes in T2 * on the calculation. Monte Carlo simulations were performed over 1000 noisy iterations for the VFA method, the SR-VFA method, and a proposed method, SR-VFA with a T2 * correction (SR-VFA-T2 *). Dynamic T1 estimates were calculated analytically for each method, with signals modeled by the steady-state spoiled gradient echo equation. The mean and standard deviation of these estimates were calculated and compared to truth, while varying repetition time (TR), baseline and dynamic T1 , echo time (TE), baseline and dynamic T2 *, flip angles, and the number of averages on baseline scans. Additionally, the variance of T1 in the SR-VFA and SR-VFA-T2 * methods was derived analytically based on the theory of propagation of errors. This equation was used to produce an inverse-variance weighted linear combination to improve T1 mapping precision in the SR-VFA-T2 * method. Flip angle sensitivity of dynamic T1 precision in the SR-VFA and SR-VFA-T2 * methods was also performed. Substantial bias can be produced by the SR-VFA method when the ratio of the T2 * decay of the dynamic signal versus that of the baseline signals deviates from 1, with a 0.01 deviation leading to approximately a 1% bias in cases of high SNR and TR ≫ T1 . This bias can be corrected by estimating the baseline and dynamic T2 * values in this ratio via multiecho measurements. The bias and precision of the SR-VFA-T2 * method, when normalized to scan time, is found to rival and sometimes improve upon the two flip angle VFA method when an inverse variance weighted linear combination is applied across its multiecho T1 maps. The analytic variance equation presented is found to be accurate within 1% relative to the Monte Carlo simulations over a broad parameter space. Flip angle ranges that maximize SR-VFA and SR-VFA-T2 *T1 precision over a broad parameter space are given, and each is defined relative to TR and T1 . Multiecho SR-VFA-T2 * T1 mapping is found in simulation and theory to be a promising alternative to the VFA method that maintains speed of the SR-VFA method with accuracy and precision similar to the VFA method.

| openaire: EC/H2020/736937/EU//M-CUBE

Background Arterial spin labeled (ASL) 4D MRA has been shown to be a valuable method for simultaneously assessing anatomic structure and dynamic filling of cerebral arteries. Beyond improved SNR and parallel imaging performance, performing these studies at ultrahigh field will benefit from longer T1s resulting in more persistent labeling allowing for improved visualization of distal vessels, especially in conditions of reduced flow. Methods Studies were performed on Siemens 7T whole body MRI scanner with a 16-channel transceiver TEM stripline array head coil driven by a series of 16, 1 kW amplifiers (CPC, Pittsburgh, PA). Non-contrast enhanced 4D ASL MRA was performed using an ECG-triggered, segmented, 3D acquisition using signal targeting with alternating radiofrequency (STAR) for spin labeling. Label and control volumes were acquired with magnetization preparation immediately after the ECG R-wave and before the segmented readout used to generate the time resolved frames (Fig. 1). Background saturation over the imaging volume was applied for both label and control volumes. Data was acquired with 26 frames over 1700 ms (two R-R intervals) with a temporal resolution of ~60 ms and an acquisition time of 12:23. To address the challenges of B1+ homogeneity and efficiency at 7T, B1+ shimming was performed based on a small flip angle, calibration scan. Two B1+ shimming strategies were compared; 1) a static approach which used aC P-like mode over the entire brain for all RF pulses in the sequence, and 2) a dynamic approach which used a separate B1+ shim optimized for efficiency for the

A highly accurate, fast three-dimensional in vivo temperature mapping method is developed using MRI water photon chemical shift. It is important to have the precise temperature distribution information during laser-tissue thermal treatment. Several methods can be used for temperature measurement including thermal couple, optical fiber sensor, and MRI (magnetic resonance imaging) methods. MRI is the only feasible method for 3D in vivo, non-invasive temperature distribution measurement for laser-tissue interaction. The water proton chemical shift method is used in 3D MRI mapping. Varies MRI parameters, such as flip angle, TE, TR, spatial resolution, and temporal repetition, were optimized for the temperature mapping. The laser radiation of 805nm wavelength and a light-absorbing dye, indocyanine green (ICG) was used for temperature elevation. The measurement was conducted using gel phantom, chicken tissue and rats. The phantom system was constructed with a dye-enhanced spherical gel embedded in uniform gel phantom, simulating a tumor within normal tissue. The normal temperature elevation within ex vivo tissue such as chicken breast can reach up to 45-50 degree C with a power density of 1.3W/cm2 (with laser power of 3W and 1.7cm beam size). The temperature resolution is 0.37 degree C with a 0.2-mm spatial resolution and repetition rate of around 40 seconds. The external magnetic field drift effect is also evaluated.

Theoretical Description, Measurement, and Correction of Localization Errors in 31P Chemical-Shift Imaging

A comparative numerical study of rotating and stationary RF coils in terms of flip angle and specific absorption rate for 7 T MRI

To eliminate the reference mirror (REF) error of the Fizeau interferometer for measuring X-ray mirrors, the reference calibration method of lateral multi-shift measurements at the hundreds-micrometer pixel level is presented. Because of the high aspect ratio of long X-ray mirrors, by shifting the surface under test (SUT) along the tangential direction with integer multiple pixels, we extend the calibration method by using the difference between multiple shifted measurements to build an augmented multi-matrix for extracting the two-dimensional (2D) absolute surface. The method can be applied to arbitrary measurement regions of the test optics, and the measurement for both the benchmark sub-aperture and calibration of the REF is accomplished in a single measuring process. Furthermore, by adjusting the shift to the millimeter scale, reference-subtracted sub-apertures can be stitched to obtain the absolute 2D map of the X-ray mirror. Experimental results show that all the 2D discrepancies reach sub-nanometer repeatability, and comparison results between the long trace profiler (LTP) and the proposed method have been performed. Therefore, the results demonstrated that the proposed method meets the requirements of X-ray mirror measurements.

Several 2D methods have been proposed to estimate WSS and OSI from PC-MRI, neglecting the longitudinal velocity gradients that typically arise in cardiovascular flow, particularly on vessel geometries whose cross section and centerline orientation strongly vary in the axial direction. Thus, the contribution of longitudinal velocity gradients remains understudied. In this work, we propose a 3D finite-element method for the quantification of WSS and OSI from 3D-CINE PC-MRI that accounts for both in-plane and longitudinal velocity gradients. We demonstrate the convergence and robustness of the method on cylindrical geometries using a synthetic phantom based on the Poiseuille flow equation. We also show that, in the presence of noise, the method is both stable and accurate. Using computational fluid dynamics simulations, we show that the proposed 3D method results in more accurate WSS estimates than those obtained from a 2D analysis not considering out-of-plane velocity gradients. Further, we conclude that for irregular geometries the accurate prediction of WSS requires the consideration of longitudinal gradients in the velocity field. Additionally, we compute 3D maps of WSS and OSI for 3D-CINE PC-MRI data sets from an aortic phantom and sixteen healthy volunteers and two patients. The OSI values show a greater dispersion than WSS, which is strongly dependent on the PC-MRI resolution. We envision that the proposed 3D method will improve the estimation of WSS and OSI from 3D-CINE PC-MRI images, allowing for more accurate estimates in vessels with pathologies that induce high longitudinal velocity gradients, such as coarctations and aneurisms.