Are multi-contrast magnetic resonance images necessary for segmenting multiple sclerosis brains? A large cohort study based on deep learning Publication date: January 2020 Source: Magnetic Resonance Imaging, Volume 65 Author(s): Ponnada A. Narayana, Ivan Coronado, Sheeba J. Sujit, Xiaojun Sun, Jerry S. Wolinsky, Refaat E. Gabr Abstract Background Magnetic resonance images with multiple contrasts or sequences are commonly used for segmenting brain tissues, including lesions, in multiple sclerosis (MS). However, acquisition of images with multiple contrasts increases the scan time and complexity of the analysis, possibly introducing factors that could compromise segmentation quality. Objective To investigate the effect of various combinations of multi-contrast images as input on the segmented volumes of gray (GM) and white matter (WM), cerebrospinal fluid (CSF), and lesions using a deep neural network. Methods U-net, a fully convolutional neural network was used to automatically segment GM, WM, CSF, and lesions in 1000 MS patients. The input to the network consisted of 15 combinations of FLAIR, T1-, T2-, and proton density-weighted images. The Dice similarity coefficient (DSC) was evaluated to assess the segmentation performance. For lesions, true positive rate (TPR) and false positive rate (FPR) were also evaluated. In addition, the effect of lesion size on lesion segmentation was investigated. Results Highest DSC was observed for all the tissue volumes, including lesions, when the input was combination of all four image contrasts. All other input combinations that included FLAIR also provided high DSC for all tissue classes. However, the quality of lesion segmentation showed strong dependence on the input images. The DSC and TPR values for inputs with the four contrast combination and FLAIR alone were very similar, but FLAIR showed a moderately higher FPR for lesion size <100 μl. For lesions smaller than 20 μl all image combinations resulted in poor performance. The segmentation quality improved with lesion size. Conclusions Best performance for segmented tissue volumes was obtained with all four image contrasts as the input, and comparable performance was attainable with FLAIR only as the input, albeit with a moderate increase in FPR for small lesions. This implies that acquisition of only FLAIR images provides satisfactory tissue segmentation. Lesion segmentation was poor for very small lesions and improved rapidly with lesion size.

An evaluation of the reproducibility of 1H-MRS GABA and GSH levels acquired in healthy volunteers with J-difference editing sequences at varying echo times Publication date: January 2020 Source: Magnetic Resonance Imaging, Volume 65 Author(s): James J. Prisciandaro, Mark Mikkelsen, Muhammad G. Saleh, Richard A.E. Edden Abstract Recent advances in J-difference-edited proton magnetic resonance spectroscopy (1H MRS) data acquisition and processing have led to the development of Hadamard Encoding and Reconstruction of MEGA-Edited Spectroscopy (HERMES) techniques, which enable the simultaneous measurement of ɣ-aminobutyric acid (GABA), the primary inhibitory amino acid neurotransmitter in the central nervous system, and of glutathione (GSH), the most abundant antioxidant in living tissue, at the commonly available magnetic field strength of 3 T. However, the reproducibility of brain levels of GABA and GSH measured across multiple scans in human subjects using HERMES remains to be established. In the present study, twelve healthy volunteers completed two consecutive HERMES scans of the dorsal anterior cingulate cortex (dACC) to assess the test-retest reproducibility of the technique for GABA and GSH measurements at TE = 80 ms. Eleven of the twelve participants additionally completed two consecutive MEGA-PRESS scans at TE = 120 ms, with editing pulses configured for GSH acquisition, to compare the reliability of GSH in the same voxel measured using the standard MEGA-PRESS at TE = 120 ms. The primary findings of study were that, 1) the coefficient of variation (CV) of measuring GABA with HERMES was 16.7%, which is in agreement with the reliability we previously reported for measuring GABA using MEGA-PRESS; and 2) the reliability of measuring GSH with MEGA-PRESS at TE = 120 ms was more than twice as high as that for measuring the antioxidant with HERMES at TE = 80 ms (CV = 7.3% vs. 19.0% respectively). These findings suggest that HERMES and MEGA-PRESS offer similar reliabilities for measuring GABA, while MEGA-PRESS at TE = 120 ms is more reliable for measuring GSH relative to HERMES at TE = 80 ms.

Generation of human thalamus atlases from 7 T data and application to intrathalamic nuclei segmentation in clinical 3 T T1-weighted images Publication date: January 2020 Source: Magnetic Resonance Imaging, Volume 65 Author(s): Yuan Liu, Pierre-François D'Haese, Allen T. Newton, Benoit M. Dawant Abstract The thalamus serves as the central relay station for the brain. It processes and relays sensory and motor signals between different subcortical regions and the cerebral cortex and it can be divided into several neuronal clusters referred to as nuclei. Each of these can possibly be subdivided into sub-nuclei. Accurate and reliable identification of thalamic nuclei is important for surgical interventions and neuroanatomical studies. This is however a challenging task because the small size of the nuclei and the lack of contrast over the thalamus region in clinically acquired images does not permit the visualization of their boundaries. A number of methods have been developed for thalamus parcellation but the vast majority of these relies on diffusion imaging or functional imaging. The low resolution of these images only permit localizing the largest nuclei. In this work we propose a method to segment smaller nuclei. We first present a protocol to build histological-like atlases from a series of high-field (7 Tesla) MR images acquired with different pulse sequences that each permits to visualize the boundaries of a subset of the nuclei. We use this protocol to scan 9 subjects and we manually delineate 23 thalamic nuclei following the Morel atlas naming convention for each of these subjects. Manual contours for the nuclei are subsequently utilized to create statistical shape models. With these data, we compare four methods for the segmentation of thalamic nuclei in 3 T images we have also acquired for the 9 subjects included in the study: (1) single atlas, (2) multi atlas, (3) statistical shape, and (4) hierarchical statistical shape in which thalamic nuclei are hierarchically fitted to the images, starting from the largest ones. Results of a leave-one-out validation study conducted on the nine image sets we have acquired show that the multi atlas approach improves upon the single atlas approach for most nuclei. Segmentations obtained with the hierarchical statistical shape model yield the highest accuracy, with dice coefficients ranging from 0.53 to 0.90, mean surface errors from 0.27 mm to 0.64 mm, and maximum surface errors from 1.31 mm to 2.52 mm for all nuclei averaged across test cases. This suggests the feasibility of using such approach for localizing thalamic substructures in clinically acquired MR volumes. It may have a direct impact on surgeries such as Deep Brain Stimulation procedures that require the implantation of stimulating electrodes in specific thalamic nuclei.

Quantitative ferumoxytol-enhanced MRI in pregnancy: A feasibility study in the nonhuman primate Publication date: January 2020 Source: Magnetic Resonance Imaging, Volume 65 Author(s): Ante Zhu, Scott B. Reeder, Kevin M. Johnson, Sydney M. Nguyen, Sean B. Fain, Ian M. Bird, Thaddeus G. Golos, Oliver Wieben, Dinesh M. Shah, Diego Hernando Abstract Objectives To assess the feasibility of ferumoxytol-enhanced MRI in pregnancy with a nonhuman primate model. Materials and methods In this prospective study, eleven pregnant rhesus macaques at day 98 ± 5 of gestation were divided into three groups, untreated control (UC) (n = 3), saline control (SC) (n = 4) and interleukin 1 beta (IL-1β) treated (IT) (n = 4), which were administered with either saline or IL-1β into the amniotic fluid. All animals were imaged at multiple time points before and after ferumoxytol administration (4 mg/kg). Longitudinal R2* and susceptibility of tissues were obtained using region-of-interest analysis and the longitudinal changes were assessed using linear mixed models and Student's t-test. Results In fetuses, a slope of 0.3 s−1/day (P = 0.008), 0.00 ppm/day (P = 0.699) and − 0.2 s−1/day (P = 0.023) was observed in liver R2*, liver susceptibility, and lung R2*, respectively. In placentas, R2* and susceptibility increased immediately after ferumoxytol administration (P < 0.001) and decreased to baseline within two days. The mean change from baseline showed no significant difference between the SC group and the IT group at all scan time points. In maternal livers, R2* increased immediately after ferumoxytol administration, further increased at one-day, and then decreased but remained elevated (P < 0.001). The mean change from baseline showed no significant difference between the SC group and the IT group at all scan time points. Conclusions This work demonstrates the feasibility of quantitative ferumoxytol-enhanced MRI to measure dynamics of ferumoxytol delivery and washout in the placenta. Stable MRI measurements indicated no evidence of iron deposition in fetal tissues of nonhuman primates after maternal ferumoxytol exposure.

Evaluation of B0-correction of relative CBF maps using tagging distance dependent Z-spectrum (TADDZ) Publication date: January 2020 Source: Magnetic Resonance Imaging, Volume 65 Author(s): Frederick C. Damen, Rong-Wen Tain, Riya Thomas, Weigo Li, Leon Tai, Kejia Cai Abstract Arterial spin labeling (ASL) MRI, based on endogenous contrast from blood water, is used in research and diagnosis of cerebral vascular conditions. However, artifacts due to imperfect imaging conditions such as B0-inhomogeneity (ΔB0) could lead to variations in the quantification of relative cerebral blood flow (CBF). In this study, we evaluate a new approach using tagging distance dependent Z-spectrum (TADDZ) data, similar to the ΔB0 corrections in the chemical exchange saturation transfer (CEST) experiments, to remove the imaging plane B0 inhomogeneity induced CBF artifacts in ASL MRI. Our results indicate that imaging-plane B0-inhomogeneity can lead to variations and errors in the relative CBF maps especially under small tagging distances. Along with an acquired B0 map, TADDZ data helps to eliminate B0-inhomogeneity induced artifacts in the resulting relative CBF maps. We demonstrated the effective use of TADDZ data to reduce variation while subjected to systematic changes in ΔB0. In addition, TADDZ corrected ASL MRI, with improved consistency, was shown to outperform conventional ASL MRI by differentiating the subtle CBF difference in Alzheimer's disease (AD) mice brains with different APOE genotypes.

Parametric linear vibration response studies for longitudinal whole-body gradient coils: Theoretical predictions based on an exact linear elastodynamic model Publication date: January 2020 Source: Magnetic Resonance Imaging, Volume 65 Author(s): Jamal Sakhr, Blaine A. Chronik Abstract Passive reduction of gradient coil (GC) cylinder vibration depends critically on a thorough knowledge of how all pertinent physical parameters affect the vibration response. In this paper, we employ a recently introduced linear elastodynamic Z-coil model to study how the displacement response of a whole-body GC cylinder (subject to exclusive excitation of its Z-coil windings) is affected by independent regularized variations in its: (i) length; (ii) radial thickness; (iii) mass density; (iv) Poisson ratio; and (v) Young modulus (stiffness). The results exhibit a rich variety of behaviors at different excitation frequencies, and in the parameter ranges of interest, the displacement response is found to be particularly sensitive to variations in cylinder geometry and mass density. The results also show that, with the exception of the stiffness, there are no optimal ranges of regularized values of the considered parameters that will reduce the displacement (and hence the vibration) of a GC cylinder at all frequencies of interest. For typical GC cylinder geometries and densities, and under the condition that only the Z-coil windings are excited, the model predicts that increasing the cylinder stiffness above 100 GPa will reduce vibration at all frequencies below 2000 Hz.

Quantitative renal function assessment of atheroembolic renal disease using view-shared compressed sensing based dynamic-contrast enhanced MR imaging: An in vivo study Publication date: January 2020 Source: Magnetic Resonance Imaging, Volume 65 Author(s): Hanjing Kong, Bin Chen, Xiaodong Zhang, Chengyan Wang, Min Yang, Li Yang, Xiaoying Wang, Jue Zhang Abstract Atheroembolic renal disease (AERD) is the major cause of renal insufficiency in the elderly, and particularly, the diagnose of AERD is often delayed and even missed due to its nonspecific presentation and the sudden occurrence of an embolic event. To investigate the feasibility of the view-shared compressed sensing (VCS) based dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) in the assessment of AERD in animal models. The reproducibility of VCS DCE-MRI based glomerular filtration rate (GFR) estimation was first evaluated using the three healthy rabbits. Animal models of unilateral AERD were then conducted. All the rabbits underwent VCS DCE-MRI and the GFR maps were estimated by a commonly used cortical-compartment model. The whole kidney and suspicious lesion region GFR values of embolized kidneys were then compared with the corresponding values of normal kidneys. Finally, the suspicious lesion regions were confirmed by the corresponding renal specimens and histological findings. The reproducibility of GFR measurements was analyzed using the coefficient of variation and Bland-Altman analysis. The GFR values of normal and embolized kidneys were compared using the Student t-test. Contrast-enhanced images with sufficient diagnostic quality and reduced motion artifacts are obtained at a temporal resolution of 2.5 s. The Bland-Altman plot indicated close agreement between the GFR values estimated from between-day scans in healthy rabbits. Besides, there existed significant differences between the pixel-wise GFR values of normal and AERD kidneys in region-based comparison(P < 0.0001). The suspicious lesions are consistent well with the renal specimen and histological findings. The preliminary animal study verified the feasibility of VCS DCE-MRI for renal function evaluation, and the strategy could potentially provide a valuable tool to identify AERD.

Differential detection of metastatic and inflammatory lymph nodes using intravoxel incoherent motion diffusion-weighted imaging Publication date: January 2020 Source: Magnetic Resonance Imaging, Volume 65 Author(s): Liuji Guo, Xiaomin Liu, Zhi Liu, Xiaodan Li, Zhiguang Si, Jie Qin, Yingjie Mei, Zhongping Zhang, Yikai Xu, Yuankui Wu Abstract Purpose This study sought to monitor the dynamic process of lymph node (LN) metastasis with intravoxel incoherent motion diffusion-weighted imaging (IVIM-DWI), and to investigate the impact of disease course on the detection of metastatic LNs by IVIM-DWI. Methods Twenty female New Zealand rabbits with 2.5–3.0 kg body weight were studied. VX2 cells and egg yolk emulsion were randomly inoculated into one thigh to induce metastatic and inflammatory popliteal LNs, respectively. Eight rabbits underwent IVIM-DWI (14 b values, 0–2000 s/mm2) 2 h prior to, and 14, 21, and 28 days after inoculation (D0, D14, D21, D28). The apparent diffusion coefficient (ADC), true diffusion coefficient (D), pseudodiffusion coefficient (D*), and perfusion fraction (f) were measured and compared between the metastatic and the inflammatory groups at each time point. Three rabbits randomly chosen from the remaining twelve rabbits were sacrificed at each time point to perform hematoxylin and eosin staining and histologic evaluation. Results The patterns of dynamic change of D*, ADC, and D were different between the metastatic and the inflammatory LNs. The metastatic group had a lower D* value at D14 (p = .003), and greater ADC and D values at both D21 (p = .001, p = .001) and D28 (p = .021, p = .001), compared to the inflammatory group. The f value of the metastatic group was greater than that of the inflammatory only at D28 (p = .001). Conclusions IVIM-DWI can reflect the dynamic process of LN metastasis, and disease course has a significant influence on the ability of IVIM-DWI to detect metastatic nodes.

Visualizing the lateral habenula using susceptibility weighted imaging and quantitative susceptibility mapping Publication date: January 2020 Source: Magnetic Resonance Imaging, Volume 65 Author(s): Naying He, Sean K. Sethi, Chencheng Zhang, Yan Li, Yongsheng Chen, Bomin Sun, Fuhua Yan, E. Mark Haacke Abstract The habenulae consist of a pair of small nuclei which bridge the limbic forebrain and midbrain monoaminergic centers. They are implicated in major depressive disorders due to abnormal phasic response when provoked by a conditioned stimulus. The lateral habenula (Lhb) is believed to be involved in dopamine metabolism and is now a target for deep brain stimulation, a treatment which has shown promising anti-depression effects. We imaged the habenulae with susceptibility weighted imaging (SWI) and quantitative susceptibility mapping (QSM) in order to localize the lateral habenula. Fifty-six healthy controls were recruited for this study. For the quantitative assessment, we traced the structure to compute volume from magnitude images and mean susceptibility bilaterally for the habenula on QSM. Thresholding methods were used to delineate the Lhb habenula on QSM. SWI, true SWI (tSWI), and QSM data were subjectively reviewed for increased Lhb contrast. SWI, QSM, and tSWI showed bilateral signal changes in the posterior location of the habenulae relative to the anterior location, which may indicate increased putative iron content within the Lhb. This signal behavior was shown in 41/44 (93%) subjects. In summary, it is possible to localize the lateral component of the habenula using SWI and QSM at 3 T.

Prognosticating brain tumor patient survival after laser thermotherapy: Comparison between neuroradiological reading and semi-quantitative analysis of MRI data Publication date: January 2020 Source: Magnetic Resonance Imaging, Volume 65 Author(s): Jonathan M. Hanna, Danielle Temares, Fahmeed Hyder, Douglas L. Rothman, Robert K. Fulbright, Veronica L. Chiang, Daniel Coman Abstract Background and purpose Given increasing interest in laser interstitial thermotherapy (LITT) to treat brain tumor patients, we explored if examining multiple MRI contrasts per brain tumor patient undergoing surgery can impact predictive accuracy of survival post-LITT. Materials and methods MRI contrasts included fluid-attenuated inversion recovery (FLAIR), T1 pre-gadolinium (T1pre), T1 post-gadolinium (T1Gd), T2, diffusion-weighted imaging (DWI), apparent diffusion coefficient (ADC), susceptibility weighted images (SWI), and magnetization-prepared rapid gradient-echo (MPRAGE). The latter was used for MRI data registration across preoperative to postoperative scans. Two ROIs were identified by thresholding preoperative FLAIR (large ROI) and T1Gd (small ROI) images. For each MRI contrast, a numerical score was assigned based on changing image intensity of both ROIs (vs. a normal ROI) from preoperative to postoperative stages. The fully-quantitative method was based on changing image intensity across scans at different stages without any human intervention, whereas the semi-quantitative method was based on subjective criteria of cumulative trends across scans at different stages. A fully-quantitative/semi-quantitative score per patient was obtained by averaging scores for each MRI contrast. A standard neuroradiological reading score per patient was obtained from radiological interpretation of MRI data. Scores from all 3 methods per patient were compared against patient survival, and re-examined for comorbidity and pathology effects. Results Patient survival correlated best with semi-quantitative scores obtained from T1Gd, ADC, and T2 data, and these correlations improved when biopsy and comorbidity were included. Conclusion These results suggest interfacing neuroradiological readings with semi-quantitative image analysis can improve predictive accuracy of patient survival.