Jue Hou¹, Courtney Bauer¹, Mary McDougall^1,2, and Steve Wright^1,2
1Electrical and Computer Engineering, Texas A&M University, College Station, TX, United States, ²Biomedical Engineering, Texas A&M University, College Station, TX, United States

23Na and 31P spectroscopy are powerful tools in assessing muscles in Duchenne Muscular Dystrophy studies. Frequency translation has been previously introduced as a means to facilitate multichannel studies on systems equipped with only 1H receiver arrays. Many approaches to frequency translation require mixing on both transmit and receive, which bypasses the need for phase correction.  Presented here is an approach to receive-only translation of acquired data to the 1H frequency, that allows for post-processing phase correction using signals coupled from the host system’s and translator’s local oscillators. 

Hanwen Liu^1,2, Vladimir Grouza^1,2, Marius Tuznik^1,2, and David Rudko^1,2,3
1McConnell Brain Imaging Centre, Montreal Neurological Institute and Hospital, Montreal, QC, Canada, ²Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada, ³Department of Biomedical Engineering, McGill University, Montreal, QC, Canada

Multi-echo gradient echo (GRE) based, myelin water imaging (MWI) is subject to data quality constraints related to signal to noise ratio (SNR), B₀ field inhomogeneities and gradient imperfections. These constraints pose challenges for conventional model-free, non-negative least squares (NNLS) fitting. As an alternative, a three-pool parametric model can be applied, with nonlinear least squares(NLLS) fitting to process MWI data. The three-pool model confers stability, but has the disadvantage of making prior assumptions. This study investigated the feasibility of using a novel, model-free approach called spectrum analysis for multiple exponentials via experimental condition oriented simulation (SAME-ECOS) for the GRE MWI analysis.

Ye Wu¹, Sahar Ahmad¹, and Pew-Thian Yap^1
1Department of Radiology and Biomedical Research Imaging Center (BRIC), University of North Carolina at Chapel Hill, Chapel Hill, NC, United States

Connectivity-based parcellation of subcortical structures typically relies on the pre-parcellation of the cortex. Here, we propose an unsupervised data-driven approach for subcortical parcellation based on diffusion tractography without relying on the annotation of cortical regions or fiber tracts.

Nicholas John Sisco¹, Elizabeth G Keeling¹, Aimee Borazanci¹, Richard D Dortch¹, and Ashley M Stokes^1
1Division of Neuroimaging Research, Barrow Neurological Institute, Phoenix, AZ, United States

In this study, we propose a generalized linear solution to estimate dynamic relaxation parameters R2* and R2 from combined spin- and gradient-echo (SAGE) MRI. We compared linear least squares (LLSQ) to nonlinear least squares (NLSQ) using Monte-Carlo simulated data and in vivo whole-brain data with varying signal-to-noise ratios (SNR). We show that using the LLSQ is both computationally more efficient and retains NLSQ precision, producing nearly the same estimates of R2* and R2 in a fraction of the time. This approach is widely extendable to other multi-echo, multi-contrast MRI methods, with applications including both perfusion MRI and functional MRI.

David C Zhu¹, Chih-Ying Gwo², An-Wen Deng², Norman Scheel¹, Mari A Dowling¹, and Rong Zhang^3
1Michigan State University, East Lansing, MI, United States, ²Chien Hsin University of Science and Technology, Taoyuan, Taiwan, ³University of Texas Southwestern Medical Center, Dallas, TX, United States

Alzheimer’s disease (AD) is a neurodegenerative disease and the most common cause of dementia among older adults. We implemented 3D Zernike transformation to characterize the shape changes of hippocampus in 428 older subjects with high-quality T₁-weighted volumetric brain scans. The hippocampus shape features characterized with 3D Zernike transformation, in complement to volume measures, may serve as a novel imaging marker to monitor clinical AD progression.

Aldo Camargo¹ and Ze Wang^2
1UMB, Baltimore, MD, United States, ²Radiology, University of Maryland of Baltimore, Baltimore, MD, United States

We proposed a novel method to compute arterial transit time from ASL MRI acquired at a single post-labeling-delay time. Using the method, we found significant difference between NC and AD subjects.

Ye Wu¹, Sahar Ahmad¹, Khoi Minh Huynh¹, Tiantian Xu¹, and Pew-Thian Yap^1
1Department of Radiology and Biomedical Research Imaging Center (BRIC), University of North Carolina at Chapel Hill, Chapel Hill, NC, United States

Processing and analyzing diffusion MRI (dMRI) data is important for uncovering the neural underpinnings of white matter (WM) development and degeneration across the human lifespan. Here, we introduce a robust and integrative toolkit, called Computational Medical Imaging (COMEDI), for processing, analyzing, and visualizing lifespan dMRI data.

Siyao Du¹, Mengfan Wang¹, Shasha Liu¹, Xiaoqian Bian¹, Xinyue Chen¹, Liangcun Guo¹, Guoliang Huang¹, Ruimeng Zhao¹, Can Peng¹, Wenhong Jiang¹, Qinglei Shi², Xu Yan², Guang Yang³, and Lina Zhang^1
1Department of Radiology, The First Affiliated Hospital of China Medical University, Shenyang, China, ²MR Scientific Marketing, Siemens Healthineers Ltd., Beijing, China, ³Shanghai Key Laboratory of Magnetic Resonance, East China Normal University, Shanghai, China

In this study, we compared first-order features of six DWI models in predicting the HER-2 expression in breast invasive ductal carcinoma, including conventional mono-exponential model derived parameter ADC, IVIM derived parameters (D, Dstar, f), DKI derived parameters (D, K), SEM derived parameters (DDC, α), FROC derived parameters (D, β, mu), CTRW derived parameters (D, α, β), and a comprehensive diagnostic logistic regression model was established to improve the diagnostic performance. We found that CTRM and IVIM derived first-order features demonstrated the powerful performance, and their combination reached the highest performance. The finding has a great value in breast cancer treatment.

Xuetong Zhou^1,2, Philip K. Lee^2,3, Daehyun Yoon², and Brian A. Hargreaves^1,2,3
1Bioengineering, Stanford University, Stanford, CA, United States, ²Radiology, Stanford University, Stanford, CA, United States, ³Electrical Engineering, Stanford University, Stanford, CA, United States

Metal implants induce severe B0 field inhomogeneities, causing severe distortion near the implant. In this work, we combined multi-echo spin echo and parallel imaging to an approach that uses echo shifting to encode spectral components and resolve the metal-induced artifacts. Phantom and in vivo experiments were conducted and indicated that the proposed method provides comparable artifact reduction and image quality to existing multispectral imaging approaches in comparable scan time. 

Mehmet Yigit Avci^1,2, Ziyu Li³, Qiuyun Fan^2,4,5, Susie Huang^2,4, Berkin Bilgic^2,4, and Qiyuan Tian^2,4
1Electrical and Electronics Engineering, Bogazici University, Istanbul, Turkey, ²Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA, United States, ³Nuffield Department of Clinical Neurosciences, University of Oxford, Oxford, United Kingdom, ⁴Department of Radiology, Harvard Medical School, Boston, MA, United States, ⁵Department of Biomedical Engineering,College of Precision Instruments and Optoelectronics Engineering, Tianjin University, Tianjin, China

Neural networks reduce the data requirement for deep learning-based quantitative MRI, nonetheless their uncertainty/confidence has rarely been characterized. We implemented Monte Carlo dropout, a Bayesian approximation of Gaussian process, using U-Net that include dropout layers (active during training and inference) to address this. The uncertainty was calculated as the variance of predictions from 100 different dropout configurations. The estimates were calculated as the average of predictions. The proposed method also achieved higher accuracy in estimating FA and MD from only 3 diffusion-weighted images compared to standard U-Net, which was readily usable for other MRI applications (reconstruction, super-resolution, denoising, segmentation, classification).

Haifeng Wang¹, Yongquan Ye², Congcong Liu¹, Jingyuan Lv², Zhilang Qiu¹, Xin Liu¹, Jian Xu², Dong Liang¹, and Hairong Zheng^1
1Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, ²UIH America, Inc., Houston, TX, United States

A novel single-scan hybrid 3D imaging method using Wave-CAIPI and MULTIPLEX technologies, named as WAMP, is proposed for 3D high-resolution rapid imaging. One single scan of the proposed rapid imaging method not only generates simultaneous B1t and T1 maps, but also qualitative images of T1W, PDW, T2*W, augmented T1W (aT1W), SWI and MRA, as well as parametric maps of T2* (R2*), PD and QSM. The in vivo experiments have showed that the proposed method could rapidly achieve high image quality and high quantification accuracy at the high acceleration factors in clinical applications.

Stephen Pickup¹, Miguel Romanello Giroud Joaquim¹, Hoon Choi¹, Mamta Gupta¹, Cynthia Clendenin¹, Hee Kwon Song¹, and Rong Zhou^1
1Department of Radiology, University of Pennsylvania, Philadelphia, PA, United States

Dynamic contrast enhanced MRI data in the abdomens of small animal models is often corrupted due to the effects respiratory and peristaltic motion.  Here a DCE protocol that employs stack of stars sampling throughout was implemented and was shown to be robust with respect to motion artifacts.  The sampling scheme also facilitates image reconstruction methods that employ view sharing, notably KWIC, yielding images with high temporal and spatial resolution.  The protocol was demonstrated in an orthotopic murine model of pancreatic cancer.  The resulting data were analyzed using the reference tissue method and provided high quality K_trans and v_e parameter maps. 

Myung-Ho In¹, Daehun Kang¹, Hang Joon Jo², Uten Yarach³, Nolan K Meyer^1,4, Joshua D Trzasko¹, John Huston III¹, Matt A Bernstein¹, and Yunhong Shu^1
1Department of Radiology, Mayo Clinic, Rochester, MN, United States, ²Department of Physiology, College of Medicine, Hanyang University, Seoul, Korea, Republic of, ³Department of Radiologic Technology, Faculty of Associated Medical Sciences, Chiang Mai University, Chiang Mai, Thailand, ⁴Mayo Clinic Graduate School of Biomedical Sciences, Mayo Clinic, Rochester, MN, United States

An interleaved, reverse gradient fMRI (RG-fMRI) with a point-spread-function mapping-based approach was recently proposed to minimize geometric distortion and signal dropout in echo-planar-imaging (EPI) by acquiring a pair of EPIs with the opposite (forward and reverse) phase-encoding gradient polarity and combining them after distortion correction. This study showed that the effective echo-shift map can be reliably obtained from PSF mapping to predict local signal dropouts. The predicted echo-shifting can provide guidance for protocol optimization for RG-fMRI using partial Fourier (PF) acquisition.

Guanhua Wang¹, Neel Shah², Keyue Zhu², Douglas C. Noll¹, and Jeffrey A. Fessler^2
1Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States, ²EECS, University of Michigan, Ann Arbor, MI, United States

MIRTorch (Michigan Image Reconstruction Toolbox for PyTorch) is an image reconstruction toolbox implemented with native Python/PyTorch. It provides fast iterative and data-driven image reconstruction on both CPUs and GPUs. Researchers can rapidly develop new model-based and learning-based methods (i.e., unrolled neural networks)  with its convenient abstraction layers. With the full support of auto-differentiation, one may optimize imaging protocols, such as sampling patterns and image reconstruction parameters with gradient-based methods.