Study of reproducibility of human arterial plaque reconstruction and its effects on stress analysis based on multispectral in vivo magnetic resonance imaging

Purpose: To quantify the uncertainties of carotid plaque morphology reconstruction based on patient-specific multispectral in vivo magnetic resonance imaging (MRI) and their impacts on the plaque stress analysis. Materials and Methods: In this study, three independent investigators were invited to reconstruct the carotid bifurcation with plaque based on MR images from two subjects to study the geometry reconstruction reproducibility. Finite element stress analyses were performed on the carotid bifurcations, as well as the models with artificially modified plaque geometries to mimic the image segmentation uncertainties, to study the impacts of the uncertainties to the stress prediction. Results: Plaque reconstruction reproducibility was generally high in the study. The uncertainties among interobservers are around one or the subpixel level. It also shows that the predicted stress is relatively less sensitive to the arterial wall segmentation uncertainties, and more affected by the accuracy of lipid region definition. For a model with lipid core region artificially increased by adding one pixel on the lipid region boundary, it will significantly increase the maximum Von Mises Stress in fibrous cap (>100%) compared with the baseline model for all subjects. Conclusion: The current in vivo MRI in the carotid plaque could provide useful and reliable information for plaque morphology. The accuracy of stress analysis based on plaque geometry is subject to MRI quality. The improved resolution/quality in plaque imaging with newly developed MRI protocols would generate more realistic stress predictions.

Carotid arterial plaque stress analysis using fluid-structure interactive simulation based on in-vivo magnetic resonance images of four patients

The rupture of atherosclerotic plaques is known to be associated with the stresses that act on or within the arterial wall. The extreme wall tensile stress (WTS) is usually recognized as a primary trigger for the rupture of vulnerable plaque. The present study used the in-vivo high-resolution multi-spectral magnetic resonance imaging (MRI) for carotid arterial plaque morphology reconstruction. Image segmentation of different plaque components was based on the multi-spectral MRI and co-registered with different sequences for the patient. Stress analysis was performed on totally four subjects with different plaque burden by fluid-structure interaction (FSI) simulations. Wall shear stress distributions are highly related to the degree of stenosis, while the level of its magnitude is much lower than the WTS in the fibrous cap. WTS is higher in the luminal wall and lower at the outer wall, with the lowest stress at the lipid region. Local stress concentrations are well confined in the thinner fibrous cap region, and usually locating in the plaque shoulder; the introduction of relative stress variation during a cycle in the fibrous cap can be a potential indicator for plaque fatigue process in the thin fibrous cap. According to stress analysis of the four subjects, a risk assessment in terms of mechanical factors could be made, which may be helpful in clinical practice. However, more subjects with patient specific analysis are desirable for plaque-stability study.

{\it In vivo\/} magnetic resonance studies of experimental liver disease: Carbon tetrachloride hepatotoxicity and alcohol-induced fatty liver in rat

Magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS) were used to non-invasively determine if cirrhosis induced by carbon tetrachloride (CCl$\sb4$) and phospholipase-D (PLD) could be distinguished from fatty infiltration in rat. MRS localization and water suppression methods were developed, implemented and evaluated in terms of their application to in vivo proton NMR studies of experimental liver disease. MRS studies were also performed to quantitate fatty infiltration resulting from carbon tetrachloride (CCl$\sb4$) or alcohol (ethanol) administration and the MRS results were confirmed using biochemical total lipid analysis and histology. $\rm T\sb1$ weighted MR images acquired weekly, 48 hours post administration, demonstrated only a slight increase in overall liver intensity with CCl$\sb4$ or alcohol administration, which is consistent with previously reported results. The MR images were able to detect nodules resulting from CCl$\sb4$+PLD induced cirrhosis as hypointense regions, also consistent with previous reports. Localized in vivo water and lipid proton $\rm T\sb1$ relaxation time measurements were performed and demonstrated no statistically significant trends for either agent. In vivo proton spectra were also acquired using stimulated echo techniques to quantitatively follow the changes in liver lipid content. The changes in liver lipid content observed using MRS were verified by total lipid analysis using the Folch technique and histology. The in vivo $\rm T\sb1$ and lipid quantification data str inconsistent with the previous hypothesis that the changes in $\rm T\sb1$ weighted images were the result of increased "free" water content and, therefore, increased water $\rm T\sb1$ relaxation times. These data indicate that the long term changes are more likely the result of changes in lipid content. The data are also shown to agree with the accepted hypothesis that the time course and mechanism of fatty infiltration are different for CCl$\sb4$ and alcohol. The hypothesis that the lipids resulting from either protocol are from the same lipid fraction(s), presumably triglycerides, is also supported. And lastly, on the basis of MR images and quantitative MRS lipid information, it was shown that cirrhosis could be distinguished from fatty infiltration. ^

In vivo magnetic resonance vascular imaging using laser-polarized 3He microbubbles

Laser-polarized gases (3He and 129Xe) are currently being used in magnetic resonance imaging as strong signal sources that can be safely introduced into the lung. Recently, researchers have been investigating other tissues using 129Xe. These studies use xenon dissolved in a carrier such as lipid vesicles or blood. Since helium is much less soluble than xenon in these materials, 3He has been used exclusively for imaging air spaces. However, considering that the signal of 3He is more than 10 times greater than that of 129Xe for presently attainable polarization levels, this work has focused on generating a method to introduce 3He into the vascular system. We addressed the low solubility issue by producing suspensions of 3He microbubbles. Here, we provide the first vascular images obtained with laser-polarized 3He. The potential increase in signal and absence of background should allow this technique to produce high-resolution angiographic images. In addition, quantitative measurements of blood flow velocity and tissue perfusion will be feasible.

Association between biomechanical structural stresses of atherosclerotic carotid plaques and subsequent ischaemic cerebrovascular events - A longitudinal in vivo magnetic resonance imaging-based finite element study

Background: High-resolution magnetic resonance (MR) imaging has been used for MR imaging-based structural stress analysis of atherosclerotic plaques. The biomechanical stress profile of stable plaques has been observed to differ from that of unstable plaques; however, the role that structural stresses play in determining plaque vulnerability remains speculative. Methods: A total of 61 patients with previous history of symptomatic carotid artery disease underwent carotid plaque MR imaging. Plaque components of the index artery such as fibrous tissue, lipid content and plaque haemorrhage (PH) were delineated and used for finite element analysis-based maximum structural stress (M-C Stress) quantification. These patients were followed up for 2 years. The clinical end point was occurrence of an ischaemic cerebrovascular event. The association of the time to the clinical end point with plaque morphology and M-C Stress was analysed. Results: During a median follow-up duration of 514 days, 20% of patients (n=12) experienced an ischaemic event in the territory of the index carotid artery. Cox regression analysis indicated that M-C Stress (hazard ratio (HR): 12.98 (95% confidence interval (CI): 1.32-26.67, pZ0.02), fibrous cap (FC) disruption (HR: 7.39 (95% CI: 1.61e33.82), p Z 0.009) and PH (HR: 5.85 (95% CI: 1.27e26.77), p Z 0.02) are associated with the development of subsequent cerebrovascular events. Plaques associated with future events had higher M-C Stress than those which had remained asymptomatic (median (interquartile range, IQR): 330 kPa (229e494) vs. 254 kPa (166-290), p Z0.04). Conclusions: High biomechanical structural stresses, in addition to FC rupture and PH, are associated with subsequent cerebrovascular events.

Neuromuscular disuse and atrophy: {\it In vivo\/} magnetic resonance studies of intramuscular lipid metabolism and fluid shift changes

We postulated that neuromuscular disuse results in deleteriously affected tissue-vascular fluid exchange processes and subsequently damages the important oxidative bioenergetic process of intramuscular lipid metabolism. The in-depth research reported in the literature is somewhat limited by the ex vivo nature and sporadic time-course characterization of disuse atrophy and recovery. Thus, an in vivo controlled, localized animal model of disuse atrophy was developed in one of the hindlimbs of laboratory rabbits (employing surgically implanted tetrodotoxin (TTX)-filled mini-osmotic pump-sciatic nerve superfusion system) and tested repeatedly with magnetic resonance (MR) throughout the 2-week period of temporarily induced disuse and during the recovery period (following explantation of the TTX-filled pump) for a period of 3 weeks. Controls consisted of saline/"sham"-implanted rabbit hindlimbs. The validity of this model was established with repeated electrophysiologic nerve conduction testing using a clinically appropriate protocol and percutaneously inserted small needle stimulating and recording electrodes. Evoked responses recorded from proximal (P) and distal (D) sites to the sciatic nerve cuff in the TTX-implanted group revealed significantly decreased (p $<$ 0.001) proximal-to-distal (P/D) amplitude ratios (as much as 50-70% below Baseline/pre-implanted and sham-implanted group values) and significantly increased (p $<$ 0.01) differential latency (PL-DL) values (as much as 1.5 times the pre- and sham-implanted groups). By Day 21 of recovery, observed P/D and PL-DL levels matched Baseline/sham-implemented levels. MRI-determined cross-sectional area (CSA) values of Baseline/pre-implanted, sham- or TTX-implanted, and recovering/explanted and the corresponding contralateral hindlimb tibialis anterior (TA) muscles normalized to tibial bone (TB) CSA (in TA/TB ratios) revealed that there was a significant decline (indicative of atrophic response) from pre- and sham-implanted controls by as much as 20% (p $<$ 0.01) at Day 7 and 50-55% (p $<$ 0.001) at Day 13 of TTX-implantation. In the non-implanted contralaterals, a significant increase (indicative of hypertrophic response) by as much as 10% (p $<$ 0.025) at Day 7 and 27% (p $<$ 0.001) at Day 13 + TTX was found. The induced atrophic/hypertrophic TA muscles were observed to be fully recovered by Day 21 post-explantation as evidenced by image TA/TB ratios. End-point biopsy results from a small group of rabbits revealed comprehensive atrophy of both Type I and Type II fibers, although the heterogeneity of the response supports the use of image-guided, volume-localized proton magnetic resonance spectroscopy (MRS) to noninvasively assess tissue-level metabolic changes. MRS-determined results of a 0.25cc volume of tissue within implanted limb TA muscles under resting/pre-ischemic, ischemic-stressed, and post-ischemic conditions at timepoints during and following disuse atrophy/recovery revealed significantly increased intramuscular spectral lipid levels, as much as 2-3 times (p $<$ 0.01) the Baseline/pre-implanted values at Day 7 and 6-7 times (p $<$ 0.001) at Day 13 + TTX, which approached normal levels (compared to pre- and sham-implanted groups) by Day 21 of post-explanation recovery. (Abstract shortened by UMI.) ^

“Improved Feature Extraction and Classification Techniques for Multispectral Brain Magnetic Resonance Images”

Magnetic Resonance Imaging (MRI) is a multi sequence medical imaging technique in which stacks of images are acquired with different tissue contrasts. Simultaneous observation and quantitative analysis of normal brain tissues and small abnormalities from these large numbers of different sequences is a great challenge in clinical applications. Multispectral MRI analysis can simplify the job considerably by combining unlimited number of available co-registered sequences in a single suite. However, poor performance of the multispectral system with conventional image classification and segmentation methods makes it inappropriate for clinical analysis. Recent works in multispectral brain MRI analysis attempted to resolve this issue by improved feature extraction approaches, such as transform based methods, fuzzy approaches, algebraic techniques and so forth. Transform based feature extraction methods like Independent Component Analysis (ICA) and its extensions have been effectively used in recent studies to improve the performance of multispectral brain MRI analysis. However, these global transforms were found to be inefficient and inconsistent in identifying less frequently occurred features like small lesions, from large amount of MR data. The present thesis focuses on the improvement in ICA based feature extraction techniques to enhance the performance of multispectral brain MRI analysis. Methods using spectral clustering and wavelet transforms are proposed to resolve the inefficiency of ICA in identifying small abnormalities, and problems due to ICA over-completeness. Effectiveness of the new methods in brain tissue classification and segmentation is confirmed by a detailed quantitative and qualitative analysis with synthetic and clinical, normal and abnormal, data. In comparison to conventional classification techniques, proposed algorithms provide better performance in classification of normal brain tissues and significant small abnormalities.

In vivo biochemical 7.0 Tesla magnetic resonance: preliminary results of dGEMRIC, zonal T2, and T2* mapping of articular cartilage

INTRODUCTION: Ultra-high-field whole-body systems (7.0 T) have a high potential for future human in vivo magnetic resonance imaging (MRI). In musculoskeletal MRI, biochemical imaging of articular cartilage may benefit, in particular. Delayed gadolinium-enhanced MRI of cartilage (dGEMRIC) and T2 mapping have shown potential at 3.0 T. Although dGEMRIC, allows the determination of the glycosaminoglycan content of articular cartilage, T2 mapping is a promising tool for the evaluation of water and collagen content. In addition, the evaluation of zonal variation, based on tissue anisotropy, provides an indicator of the nature of cartilage ie, hyaline or hyaline-like articular cartilage.Thus, the aim of our study was to show the feasibility of in vivo dGEMRIC, and T2 and T2* relaxation measurements, at 7.0 T MRI; and to evaluate the potential of T2 and T2* measurements in an initial patient study after matrix-associated autologous chondrocyte transplantation (MACT) in the knee. MATERIALS AND METHODS: MRI was performed on a whole-body 7.0 T MR scanner using a dedicated circular polarization knee coil. The protocol consisted of an inversion recovery sequence for dGEMRIC, a multiecho spin-echo sequence for standard T2 mapping, a gradient-echo sequence for T2* mapping and a morphologic PD SPACE sequence. Twelve healthy volunteers (mean age, 26.7 +/- 3.4 years) and 4 patients (mean age, 38.0 +/- 14.0 years) were enrolled 29.5 +/- 15.1 months after MACT. For dGEMRIC, 5 healthy volunteers (mean age, 32.4 +/- 11.2 years) were included. T1 maps were calculated using a nonlinear, 2-parameter, least squares fit analysis. Using a region-of-interest analysis, mean cartilage relaxation rate was determined as T1 (0) for precontrast measurements and T1 (Gd) for postcontrast gadopentate dimeglumine [Gd-DTPA(2-)] measurements. T2 and T2* maps were obtained using a pixelwise, monoexponential, non-negative least squares fit analysis; region-of-interest analysis was carried out for deep and superficial cartilage aspects. Statistical evaluation was performed by analyses of variance. RESULTS: Mean T1 (dGEMRIC) values for healthy volunteers showed slightly different results for femoral [T1 (0): 1259 +/- 277 ms; T1 (Gd): 683 +/- 141 ms] compared with tibial cartilage [T1 (0): 1093 +/- 281 ms; T1 (Gd): 769 +/- 150 ms]. Global mean T2 relaxation for healthy volunteers showed comparable results for femoral (T2: 56.3 +/- 15.2 ms; T2*: 19.7 +/- 6.4 ms) and patellar (T2: 54.6 +/- 13.0 ms; T2*: 19.6 +/- 5.2 ms) cartilage, but lower values for tibial cartilage (T2: 43.6 +/- 8.5 ms; T2*: 16.6 +/- 5.6 ms). All healthy cartilage sites showed a significant increase from deep to superficial cartilage (P < 0.001). Within healthy cartilage sites in MACT patients, adequate values could be found for T2 (56.6 +/- 13.2 ms) and T2* (18.6 +/- 5.3 ms), which also showed a significant stratification. Within cartilage repair tissue, global mean values showed no difference, with 55.9 +/- 4.9 ms for T2 and 16.2 +/- 6.3 ms for T2*. However, zonal assessment showed only a slight and not significant increase from deep to superficial cartilage (T2: P = 0.174; T2*: P = 0.150). CONCLUSION: In vivo T1 dGEMRIC assessment in healthy cartilage, and T2 and T2* mapping in healthy and reparative articular cartilage, seems to be possible at 7.0 T MRI. For T2 and T2*, zonal variation of articular cartilage could also be evaluated at 7.0 T. This zonal assessment of deep and superficial cartilage aspects shows promising results for the differentiation of healthy and affected articular cartilage. In future studies, optimized protocol selection, and sophisticated coil technology, together with increased signal at ultra-high-field MRI, may lead to advanced biochemical cartilage imaging.

Segmentation of Brain Tissue from Magnetic Resonance Images

Segmentation of medical imagery is a challenging problem due to the complexity of the images, as well as to the absence of models of the anatomy that fully capture the possible deformations in each structure. Brain tissue is a particularly complex structure, and its segmentation is an important step for studies in temporal change detection of morphology, as well as for 3D visualization in surgical planning. In this paper, we present a method for segmentation of brain tissue from magnetic resonance images that is a combination of three existing techniques from the Computer Vision literature: EM segmentation, binary morphology, and active contour models. Each of these techniques has been customized for the problem of brain tissue segmentation in a way that the resultant method is more robust than its components. Finally, we present the results of a parallel implementation of this method on IBM's supercomputer Power Visualization System for a database of 20 brain scans each with 256x256x124 voxels and validate those against segmentations generated by neuroanatomy experts.

Automated Segmentation and Non-Rigid Registration for the Quantitative Evaluation of Myocardial Perfusion from Magnetic Resonance Images

Myocardial perfusion quantification by means of Contrast-Enhanced Cardiac Magnetic Resonance images relies on time consuming frame-by-frame manual tracing of regions of interest. In this Thesis, a novel automated technique for myocardial segmentation and non-rigid registration as a basis for perfusion quantification is presented. The proposed technique is based on three steps: reference frame selection, myocardial segmentation and non-rigid registration. In the first step, the reference frame in which both endo- and epicardial segmentation will be performed is chosen. Endocardial segmentation is achieved by means of a statistical region-based level-set technique followed by a curvature-based regularization motion. Epicardial segmentation is achieved by means of an edge-based level-set technique followed again by a regularization motion. To take into account the changes in position, size and shape of myocardium throughout the sequence due to out of plane respiratory motion, a non-rigid registration algorithm is required. The proposed non-rigid registration scheme consists in a novel multiscale extension of the normalized cross-correlation algorithm in combination with level-set methods. The myocardium is then divided into standard segments. Contrast enhancement curves are computed measuring the mean pixel intensity of each segment over time, and perfusion indices are extracted from each curve. The overall approach has been tested on synthetic and real datasets. For validation purposes, the sequences have been manually traced by an experienced interpreter, and contrast enhancement curves as well as perfusion indices have been computed. Comparisons between automatically extracted and manually obtained contours and enhancement curves showed high inter-technique agreement. Comparisons of perfusion indices computed using both approaches against quantitative coronary angiography and visual interpretation demonstrated that the two technique have similar diagnostic accuracy. In conclusion, the proposed technique allows fast, automated and accurate measurement of intra-myocardial contrast dynamics, and may thus address the strong clinical need for quantitative evaluation of myocardial perfusion.

COVARIATE-ADJUSTED NONPARAMETRIC ANALYSIS OF MAGNETIC RESONANCE IMAGES USING MARKOV CHAIN MONTE CARLO

Permutation tests are useful for drawing inferences from imaging data because of their flexibility and ability to capture features of the brain that are difficult to capture parametrically. However, most implementations of permutation tests ignore important confounding covariates. To employ covariate control in a nonparametric setting we have developed a Markov chain Monte Carlo (MCMC) algorithm for conditional permutation testing using propensity scores. We present the first use of this methodology for imaging data. Our MCMC algorithm is an extension of algorithms developed to approximate exact conditional probabilities in contingency tables, logit, and log-linear models. An application of our non-parametric method to remove potential bias due to the observed covariates is presented.

Visibility of vascular phenylalanine in dynamic uptake studies in humans using magnetic resonance spectroscopy

In general, vascular contributions to the in vivo magnetic resonance (MR) brain spectrum are too small to be relevant. In cerebral uptake studies, however, vascular contributions may constitute a major confounder. MR visibility of vascular Phe was investigated by recording localized spectra from fully oxygenated and well-mixed whole blood. Blood Phe levels determined by MR spectroscopy (MRS) and ion-exchange chromatography showed excellent correlation. In addition, effects of blood flow were shown to have a small effect on signal amplitude with the MRS methodology used. Hence, blood Phe is almost completely MR visible at 1.5 T, even though it is severely broadened at higher fields. Without appropriate correction, cerebral Phe influx in studies of brain Phe uptake in phenylketonuria patients or healthy subjects would appear to be faster and lead to higher levels. Similar effects are envisaged for studies of ethanol or glucose uptake across the blood-brain barrier.

Generalized fuzzy clustering for segmentation of multi-spectral magnetic resonance images.

An integrated approach for multi-spectral segmentation of MR images is presented. This method is based on the fuzzy c-means (FCM) and includes bias field correction and contextual constraints over spatial intensity distribution and accounts for the non-spherical cluster's shape in the feature space. The bias field is modeled as a linear combination of smooth polynomial basis functions for fast computation in the clustering iterations. Regularization terms for the neighborhood continuity of intensity are added into the FCM cost functions. To reduce the computational complexity, the contextual regularizations are separated from the clustering iterations. Since the feature space is not isotropic, distance measure adopted in Gustafson-Kessel (G-K) algorithm is used instead of the Euclidean distance, to account for the non-spherical shape of the clusters in the feature space. These algorithms are quantitatively evaluated on MR brain images using the similarity measures.

Improved cerebellar tissue classification on magnetic resonance images of brain.

PURPOSE: To develop and implement a method for improved cerebellar tissue classification on the MRI of brain by automatically isolating the cerebellum prior to segmentation. MATERIALS AND METHODS: Dual fast spin echo (FSE) and fluid attenuation inversion recovery (FLAIR) images were acquired on 18 normal volunteers on a 3 T Philips scanner. The cerebellum was isolated from the rest of the brain using a symmetric inverse consistent nonlinear registration of individual brain with the parcellated template. The cerebellum was then separated by masking the anatomical image with individual FLAIR images. Tissues in both the cerebellum and rest of the brain were separately classified using hidden Markov random field (HMRF), a parametric method, and then combined to obtain tissue classification of the whole brain. The proposed method for tissue classification on real MR brain images was evaluated subjectively by two experts. The segmentation results on Brainweb images with varying noise and intensity nonuniformity levels were quantitatively compared with the ground truth by computing the Dice similarity indices. RESULTS: The proposed method significantly improved the cerebellar tissue classification on all normal volunteers included in this study without compromising the classification in remaining part of the brain. The average similarity indices for gray matter (GM) and white matter (WM) in the cerebellum are 89.81 (+/-2.34) and 93.04 (+/-2.41), demonstrating excellent performance of the proposed methodology. CONCLUSION: The proposed method significantly improved tissue classification in the cerebellum. The GM was overestimated when segmentation was performed on the whole brain as a single object.