Diwei Shi¹, Sisi Li², Yishi Wang³, Li Chen¹, Xiaoyu Jiang^4,5, Junzhong Xu^4,5,6,7, Quanshui Zheng¹, and Hua Guo^2
1Center for Nano and Micro Mechanics, Department of Engineering Mechanics, Tsinghua University, Beijing, China, ²Center for Biomedical Imaging Research, Tsinghua University, Beijing, China, ³Philips Healthcare, Beijing, China, ⁴Institute of Imaging Science, Vanderbilt University Medical Center, Nashville, TN, United States, ⁵Department of Radiology and Radiological Sciences, Vanderbilt University Medical Center, Nashville, TN, United States, ⁶Department of Physics and Astronomy, Vanderbilt University, Nashville, TN, United States, ⁷Department of Biomedical Engineering, Vanderbilt University, Nashville, TN, United States

Cellular microstructural parameters mapping can provide quantitative information for tumor diagnosis and treatment monitoring. Particularly, voxel-wise cell size distribution may provide critical clinical biomarkers to characterize heterogeneity of tumors. Currently, MRI-cytometry may measure such a distribution, but it “blurs” the distribution peaks and prevents differentiating different cell populations with different cell sizes. In this work, we develop a new approach to improve this and hence make it more practical to distinguish different cells. Detection of T cell infiltration for assessment of early response to immunotherapy may be a potential application of this method.

 

Sisi Li¹, Yishi Wang², Yajing Zhang³, and Hua Guo^1
1Center for Biomedical Imaging Research, Tsinghua University, Beijing, China, ²Philips Healthcare, Beijing, China, ³MR Clinical Science, Philips Healthcare, Suzhou, China

Diffusion-weighted imaging (DWI) near metal remains a technical challenge although several 3D multispectral imaging methods have achieved effective metallic artifacts correction in anatomical MRI. Additionally, considering scan time and scan flexibility, 2D multi-slice acquisition approaches are typically used in DWI. In this study, we adopted 2D Point-Spread-Function Encoded EPI to generate a series of echo-shifted in-plane distortion-corrected images and introduced an off-resonance frequency-informed reconstruction framework to further reduce metallic artifacts in DWI by through-slice signal registration. The efficacy of this method was validated in phantom experiments, in-vivo brain DTI and spinal cord DWI.

Simin Liu¹, Yunhao Xie², Hai Luo², Yishi Wang¹, Sisi Li¹, Jieying Zhang¹, Ziyi Pan¹, Ziyue Wu², and Hua Guo^1
1Center for Biomedical Imaging Research, Department of Biomedical Engineering, School of Medicine, Tsinghua University, Beijing, China, ²Marvel Stone Healthcare Co., Ltd., Wuxi, China

Low-field MR systems with state-of-the-art hardware may improve the accessibility of MRI. DWI is valuable for diagnosing diseases, while it requires multiple averages at low-field MRI considering SNR. Given similar acquisition time, multi-shot EPI DWI can be considered alternatively. Particularly, the tilted-CAIPI accelerated multi-shot PSF-EPI has been proposed in recent years to achieve fast distortion-free DWI. This study tested the feasibility of PSF-EPI on a high-performance 0.5T MR scanner, and proposed a method for further noise reduction using B0 field information. The results show that PSF-EPI eliminates distortions and improves structure fidelity at tissue boundaries compared with interleaved EPI.

Yasmin Blunck^1,2, Daniel Staeb³, Rebecca Glarin¹, Leigh Johnston^1,2, Jin Jin⁴, and Kieran O'Brien^4
1Melbourne Brain Centre Imaging Unit, The University of Melbourne, Melbourne, Australia, ²Department of Biomedical Engineering, The University of Melbourne, Melbourne, Australia, ³Siemens Healthcare Pty Ltd, Melbourne, Australia, ⁴Siemens Healthcare Pty Ltd, Brisbane, Australia

The application of diffusion weighted imaging (DWI) at ultra-high field is hampered by severe B1 inhomogeneities leading to signal dropouts in certain parts of the brain. Utilising TR-FOCI pulses in a twice-refocussed spin echo DWI sequence can effectively recover signal in these areas, albeit at the cost of increased SAR. Maintaining this enhanced B1 homogeneity, this work demonstrates a ~30% reduction in SAR based on a slice-by-slice RF amplitude optimization.

Lu Wang¹, Tian Li², Jing Cai², Liyuan Liang¹, and Hing-Chiu Chang^3
1Department of Diagnostic Radiology, University of Hong Kong, Hong Kong, China, ²Department of Health Technology and Informatics, The Hong Kong Polytechnic University, Hong Kong, China, ³Department of Biomedical Engineering, Chinese University of Hong Kong, Hong Kong, China

Four-dimensional diffusion-weighted imaging (4D-DWI) attracts lots of attention in MRI-guided radiotherapy since it can provide functional information and better delineate the tumor. However, the previously reported 4D-DW-Propeller-EPI technique used amplitude data binning with continuous interval to sort the blade data that might lead to non-uniform distribution of k-space sampling, therefore being suboptimal for subsequent Propeller-EPI reconstruction. Thus, we propose a new (K-B-optimized) binning method to consider the requirement of propeller reconstruction during the sorting of blade data, thereby improving the k-space distribution. Simulation and in-vivo studies demonstrated that the new binning method can improve the data quality of 4D-DW-Propeller-EPI.

Hinako Oshiro^1,2, Junichi Hata^1,2,3,4, Daisuke Nakashima³, Yawara Haga^1,2, Naoya Hayashi^1,2, Daisuke Yoshimaru^2,3,4, Kei Hagiya², and Hideyuki Okano^2,3
1Tokyo Metropolitan University, Tokyo, Japan, ²RIKEN Center for Brain Science, Saitama, Japan, ³Keio University School of Medicine, Tokyo, Japan, ⁴The Jikei University School of Medicine, Tokyo, Japan

We investigated the characteristics of diffusion quantification values in a restricted structure with a capillary phantom in PGSTE and OGSE. Smaller structure sizes require shorter diffusion times for structure estimation. In OGSE, we observed a transition of diffusion coefficients under 10 ms for capillary sizes from 6-12 μm, but no differences for those from 25-100 μm. Moreover, it was necessary to consider the temperature dependence of the diffusion measurement at low temperatures. We developed a system to measure the effects on diffusion quantification values at three phases of different temperatures, structure sizes, and diffusion times.

Viktor Vegh^1,2, Thomas Barrick³, Qianqian Yang⁴, Qiang Yu¹, and Martijn Cloos^1,2
1Centre for Advanced Imaging, University of Queensland, Brisbane, Australia, ²ARC Training Centre for Innovation in Biomedical Imaging Technology, Brisbane, Australia, ³Neuroscience Research Centre, St George's, University of London, London, United Kingdom, ⁴School of Mathematical Sciences, Queensland University of Technology, Brisbane, Australia

Diffusion MRI provides opportunities in probing tissue microstructure based on mm-scale measurements. Generally, an MRI diffusion model is used to infer biologically meaningful information from diffusion MRI measurements. Recently, there has been increasing interest in mapping the so-called dot-compartment in the human brain. The dot-compartment has been ascribed to a tissue compartment which has highly restricted diffusion. Using a custom diffusion MRI line scan we investigated the ability to obtain high b-value signals using 3T MRI equipped with 80mT/m gradients. It appears that the dot-compartment is elusive when probed using clinical standard scanners even when b-values of 15000s/mm² are achieved.

Zhongyan Xiong¹, Zhijun Geng², Shanshan Lian², Shaohan Yin², Guixiao Xu², Yunfei Zhang³, Yongming Dai³, Jing Zhao², Lidi Ma², Xin Liu¹, Hairong Zheng¹, Chuanmiao Xie², and Chao Zou^1
1Paul C. Lauterbur Centre for Biomedical Imaging, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China, ²Department of Radiology, Sun Yat-sen University Cancer Center, State Key Laboratory of Oncology in Southern China, Guangzhou, China, ³Central Research Institute, United Imaging Healthcare, Shanghai, China

Restriction spectrum imaging (RSI) is a novel diffusion model that captures the distinct diffusion behavior of tumors. It separates water diffusion into several microscopic compartments. The restricted compartment correlating to the tumor cellularity is expected to be a potential indicator of rectal cancer aggressiveness. To assess the ability of RSI model for rectal tumor grading, we applied a three-compartment RSI model to DWI images of patients with different histopathological grades of rectal cancer. The RSI model demonstrated its ability to discriminate the rectal cancer of low and high grades, and the results outperforms the traditional ADC model and DKI model.

Kouhei Kamiya^1,2,3, Issei Fukunaga¹, Syo Murata^1,4, Tomoko Maekawa¹, Shimpei Kato^1,3, Katsutoshi Murata⁵, Thorsten Feiweier⁶, Koji Kamagata¹, Masaaki Hori^1,2, and Shigeki Aoki^1
1Department of Radiology, Juntendo University, Tokyo, Japan, ²Department of Radiology, Toho University, Tokyo, Japan, ³Department of Radiology, The University of Tokyo, Tokyo, Japan, ⁴Department of Radiological Sciences, Komazawa University, Tokyo, Japan, ⁵Siemens Healthcare K.K., Tokyo, Japan, ⁶Siemens Healthcare GmbH, Erlangen, Germany

The ability of double diffusion encoding to estimate tissue microscopic anisotropy has gained increasing attention in clinical studies. However, the estimation of smaller values of microscopic anisotropy is known to be less precise, posing a challenge for clinical translation because many tissues of interest including the gray matter exhibit smaller values. In this study, we adopted the recently proposed denoising strategy where Marcenko‐Pastur principal component analysis (MPPCA) is applied to coil data before GRAPPA reconstruction. The results suggested this strategy is effective for improving the scan-rescan repeatability of microscopic anisotropy in the gray matter.

Diwei Shi¹, Sisi Li², Li Chen¹, Quanshui Zheng¹, Hua Guo², and Junzhong Xu^3,4,5,6
1Center for Nano and Micro Mechanics, Department of Engineering Mechanics, Tsinghua University, Beijing, China, ²Center for Biomedical Imaging Research, Tsinghua University, Beijing, China, ³Institute of Imaging Science, Vanderbilt University Medical Center, Nashville, TN, United States, ⁴Vanderbilt University Medical Center, Nashville, TN, United States, ⁵Department of Physics and Astronomy, Vanderbilt University, Nashville, TN, United States, ⁶Department of Biomedical Engineering, Vanderbilt University, Nashville, TN, United States

Quantitative microstructural imaging based on diffusion MRI usually replies on some simple gradient waveforms, with which analytical expressions can be derived e.g., for fitting cell size. However, it is challenging for this approach for modified irregular gradient waveforms that are increasingly used. Inspired by Callaghan’s matrix formalism, we propose an efficient approach for signal computation with arbitrary gradient waveforms. It can accelerate computation by three orders of magnitude with maintained accuracy, making it feasible in practical data fittings. This work paves the way for quantitative microstructural imaging with arbitrary diffusion gradient waveforms in practice. 

 

 

Yoshitaka Masutani¹ and Koh Sasaki^1,2
1Graduate School of Information Sciences, Hiroshima City University, Hiroshima, Japan, ²Hiroshima Heiwa Clinic, Hiroshima, Japan

Free water imaging (FWI) belongs to the dMRI family, and is an extention of the DTI model by adding the isotropic diffusion compartment. Conventionally, FWI parameters have been obtained by numerical fitting to measured signal values of DWI of single-shell or multi-shell.  It has been reported that it is harder to obtain robust results in single shell data. Recently, machine learning techniques have shown promising results in dMRI parameter inference. In this study, we aimed at FWI parameter inference from single-shall dMRI data by using synthetic Q-space learning. Several validation experiments by quantitative and visual assessments were performed.

Sisi Li¹, Diwei Shi², Li Chen², Quanshui Zheng², Hua Guo¹, and Junzhong Xu^3,4,5,6
1Center for Biomedical Imaging Research, Tsinghua University, Beijing, China, ²Center for Nano and Micro Mechanics, Department of Mechanics Engineering, Tsinghua University, Beijing, China, ³Institute of Imaging Science, Vanderbilt University Medical Center, Nashville, TN, United States, ⁴Department of Radiology and Radiological Sciences, Vanderbilt University Medical Center, Nashville, TN, United States, ⁵Department of Biomedical Engineering, Vanderbilt University, Nashville, TN, United States, ⁶Department of Physics and Astronomy, Vanderbilt University, Nashville, TN, United States

Quantitative microstructural imaging based on diffusion MRI provides a non-invasive means to measure microstructural parameters, e.g., cell size. However, due to the remarkable complexity of incorporating transcytolemmal water exchange into diffusion biophysical models, water exchange is usually ignored. This leads to biased estimation of microstructural parameters, particularly intracellular volume fraction. Here, we propose a new approach to incorporate water exchange naturally in diffusion biophysical models with arbitrary gradient waveforms, making it possible to fit cell size, density, and intracellular water lifetime simultaneously. This establishes a general framework to incorporate transcytolemmal water exchange in any diffusion MRI-based microstructural imaging models. 

 

 

Zuojun Wang¹, Henry Ka-Fung Mak¹, and Peng Cao^1
1Department of Diagnostic Radiology, The University of HongKong, Hong Kong, China

We develop a deep-learning-regularized single-step QSM quantification to generate QSM directly from the total phase map. A deep-learning-regularized dipole inversion network, named POCSnet, was deployed to a single-step QSM (SS-POCSnet) network, which combined a variable-SHARP (VSHARP) and the POCSnet. Meanwhile, SS-POCSnet showed improved accuracy compared with conventional single-step QSM methods. We also demonstrated the generalizability of SS-POCSnet on different datasets in vivo.

Zhuang Xiong¹, Yang Gao¹, Steffen Bollmann¹, and Hongfu Sun^1
1School of ITEE, the University of Queensland, Brisbane, Australia

Due to the intrinsic data-driven property, many existing deep learning QSM methods can only be applied to local field maps with FOV orientation and image resolution consistent with the training data. This work proposes a novel and robust deep learning approach to reconstruct QSM of arbitrary head orientation and image resolution. Experiments are conducted on both simulated and in vivo human brain data to verify the proposed approach.

Junye Yao¹, Zihan Zhou¹, Benjamin C. Tendler², Karla L. Miller², Hao Lei³, Lei Zhang^4,5, Aimin Bao^4,5, Jianhui Zhong^1,6, and Hongjian He^1
1Center for Brain Imaging Science and Technology, College of Biomedical Engineering and Instrumental, Zhejiang University, Hangzhou, China, ²Wellcome Centre for Integrative Neuroimaging, FMRIB, Nuffield Department of Clinical Neurosciences, University of Oxford, London, United Kingdom, ³State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, Innovation Academy for Precision Measurement Science and Technology, Chinese Academy of Sciences, Wuhan, China, ⁴Department of Neurology in Second Affiliated Hospital, Key Laboratory of Medical Neurobiology of Zhejiang Province, and Department of Neurobiology, Zhejiang University, Hangzhou, China, ⁵National Human Brain Bank for Health and Disease, School of Brain Science and Brain Medicine, Zhejiang University, Hangzhou, China, ⁶Department of Imaging Sciences, University of Rochester, Rochester, NY, United States

The microstructural basis of b-value dependence on FA was investigated in the human brain, using in vivo MRI and a formalin-fixed human brain hemisphere. Consistent decreases of FA with increasing b-values were observed both in vivo and ex vivo. The percentage difference of FA (pdFA) between high and low b-value datasets was found to increase with neurite density estimated from NODDI, which was further validated using histology and a multi-compartment model simulation. Results indicate that the b-value dependence of FA can be explained by compartmental differences between intra- and extra-neurites.