Suixing Zhong¹, Ya Zhang¹, Yingying Ding¹, Lisha Nie², Jing Tan¹, Conghui Ai¹, Yan Jin¹, Hongbo Wang¹, Huimei Zhang¹, Miaomiao Li¹, Rong Zhu¹, and Shangwei Gu^1
1Department of Radiology, Yunnan Cancer Hospital, Third Affiliated Hospital of Kunming Medical University, Kunming, China, ²MR Research, GE Healthcare, Beijing, China

This study aims to investigate the value of quantitative relaxation time derived from synthetic MRI (SyMRI) in the diagnosis and grading of prostate cancer (PCa). The results of this study showed that the diagnostic efficiencies of SyMRI in distinguishing PCa from most benign prostate lesions and different Gleason grades of PCa were similar to that of DWI, and the combined index of SyMRI and DWI can obtain higher diagnostic efficiencies. In addition, SyMRI has the advantages of short scanning time, unified parameters and simple operation, so it has a good prospect of clinical application.

Yi-Cheng Hsu¹, Patrick Liebig², and Ying-Hua Chu^1
1Siemens Healthineers Ltd., Shanghai, China, ²Siemens Healthcare GmbH, Erlangen, Germany

We proposed a method to simultaneously acquire T1, T2, PD weighted, and MT weighted images. The scan time is the same as conventional SPGR T1 and DESS T2 mapping but with higher accuracy. This method will benefit diagnosis on muscular dystrophy, inflammation, fibrosis, and cartilage integrity.

lidi ma¹, shanshan lian¹, huimin liu¹, tiebao meng¹, weilong zeng¹, rui zhong¹, linchang zhong¹, long qian², and chuanmiao xie^1
1Sun Yat-sen University Cancer Center, guangzhou, China, ²MR Research, GE Healthcare, Beijing, China, guangzhou, China

The clinical significance of synthetic MRI in rectal adenocarcinoma remains unclear. This study aimed to explore the of quantitative parameters derived from SyMRI clinical stage according to “DISTANCE” criteria and differentiation of rectal adenocarcinoma. Our preliminary study demonstrated that quantitative T2 and PD values obtained by SyMRI might be used for noninvasive evaluation of prognostic factors of rectal adenocarcinoma. Furthermore, combining the two quantitative relaxation metrics further improved their diagnostic performance of mrN stage in rectal adenocarcinoma.

Chenhui Li¹, Liling Long¹, Huiting Zhang², and Dominik Nickel^3
1The First Affiliated Hospital of Guangxi Medical University, Nanning, China, ²MR Scientific Marketing, Siemens Healthineers, Wuhan, China, ³MR Application Predevelopment, Siemens Healthcare GmbH, Erlangen, Germany

Two different T1 mapping sequences, based on the inversion recovery SNAPSHOT-FLASH sequence (LLIR-T1) and on B1 inhomogeneity-corrected variable flip angle (VFA-T1) acquisitions, were compared in a clinical setting. Results showed that the LLIR-T1 method had better image quality, and higher contrast between tumor and healthy liver tissue.c had a strong correlation with VFA-T1 method whether at different time point or regions. But LLIR-T1 had lower T1 in healthy liver lobe and higher T1 in tumor compared with VFA-T1.

Yuqi Yang¹, Baolong Wang¹, and Xin Zhou^1
1Key Laboratory of Magnetic Resonance in Biological Systems, State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, National Center for Magnetic Resonance in Wuhan, Wuhan Institute of Physics and Mathematics, Innovation Academy for Precision Measurement Science and Technology, Chinese Academy of Sciences–Wuhan National Laboratory for Optoelectronics, Wuhan, China

A versatile nano-vehicle, quantum dots nano-transformers, smartly transforms its shape in acidic tumor microenvironment to prolong the tumor retention, and then disassembles and releases the encapsulated photosensitizers to “turn on” the imaging contrast. The total drug dose in repeated photodynamic therapy is reduced because one injection is enough for four times of light irradiation.

Runke Wang¹, Yu Chen¹, Suhao Qiu¹, and Yuan Feng^1
1School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China

Algorithms providing accurate estimation of shear modulus and identification of tissue boundary are always desired for magnetic resonance elastography (MRE). In this study, we proposed a model-based neural network (MNN) embedding classic direct inversion (DI) and local frequency estimation (LFE). Convolution layers with Inception-like structure were applied for preprocessing, while postprocessing was implemented with a U-net structure. Additionally, DI and LFE algorithms were encapsulated as layers transforming wave images to shear modulus maps. The network performance was tested using a phantom with an inclusion. Compared with conventional algorithms, MNN provided modulus estimation with 5-fold higher contrast-to-noise ratio with clear boundaries. 

Shengyuan Ma¹ and Yuan Feng^1
1School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China

We proposed a traveling-wave expansion based subspace search (TWESS) algorithm for modulus estimation of MRE. Based on a traveling wave decomposition model, we explore the orthogonal characteristics of the signal subspace and the noise subspace. Eigenvalue decomposition of the covariance matrix of the wave data were used to search the wavenumber and retrieve the local modulus. Results showed the proposed method could provide better delineation of the inclusion boundaries, even when signal-to-noise ratio of the image was low.

Takanori Aoki¹ and Mikio Suga^1,2
1Graduate School of Science and Engineering, Chiba University, Chiba, Japan, ²Center for frontier Medical Engineering, Chiba University, Chiba, Japan

MRE is a method for non-invasively and quantitatively estimating the stiffness of biological tissues. Although many estimation methods have been proposed, such as LFE, DI, and MMDI, there is a problem that both quantitativeness and spatial resolution cannot be achieved. In this study, we developed an elastic modulus estimation method by deep learning and compared its robustness to noise, quantitativeness, and spatial resolution with existing methods by numerical simulation. Compared with the conventional method (LFE), the proposed method has the same level of noise robustness, higher spatial resolution, and better quantitative performance in the case of high elastic modulus.

Suhao Qiu¹, Linghan Kong¹, Runke Wang¹, and Yuan Feng^1
1School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China

The relationship between mechanical properties and microscopic structures of brain is of interest to both neuroscience and clinical communities. Using magnetic resonance elastography (MRE), we proposed and verified an empirical model to establish a linear correlation between shifted apparent diffusion coefficient (sADC) and shear modulus. A total of 43 healthy volunteers were included for MRE and diffusion weighted imaging (DWI) scanning. Results demonstrated that, in the parietal lobe, a strong correlation exists between shear modulus and sADC.

Xumei Hu¹, Ruokun Li², Jiahao Zhou², Jing Guo³, Ingolf Sack³, Weibo Chen⁴, He Wang⁵, Fuhua Yan², and Chengyan Wang^1
1Human Phenome Institute,Fudan University, Shanghai, China, ²Department of Radiology, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China, ³Department of Radiology, Charité–Universitätsmedizin Berlin, Berlin, Germany, ⁴Philips Healthcare, Shanghai, China, ⁵Institute of Science and Technology for Brain-inspired Intelligence, Fudan University, Shanghai, China

In this study, a combined deep learning and radiomics (DLR) approach using six different network architectures was tested and compared for the prediction of high Ki-67 expressions in patients with hepatocellular carcinoma (HCC). The model was based primarily on data from MRI and tomoelastography, a multifrequency MR elastography technique. Xception delivered the best performance and recognized seven prominent features among which four were obtained from tomoelastography. Our findings demonstrated that biomechanical properties, especially viscosity and the fluid behavior of the tumor, are crucial imaging features that are important for imaging-based cancer diagnostics.

Xinran Chen¹, Wei Wang¹, Jianpan Huang², Jian Wu¹, Lin Chen¹, Congbo Cai¹, Shuhui Cai¹, and Zhong Chen^1
1Department of Electronic Science, Xiamen University, Xiamen, China, ²Department of Biomedical Engineering, City University of Hong Kong, Hong Kong, China

Water-fat separation is a powerful tool in diagnosing many diseases and many efforts have been made to reduce the scan time. Spatiotemporally encoded (SPEN) single-shot MRI, as an emerging ultrafast MRI method, can accomplish the fastest water-fat separation since only one shot is required. However, the SPEN water/fat images  obtained by the state-of-the art methods still have some shortcomings. Here, a deep learning approach based on U-Net was proposed to obtain SPEN water/fat images simultaneously with improved spatial resolution, better fidelity and reduced reconstruction time. The efficiency of our method is demonstrated by numerical simulations, and in vivo rat experiments.

Lixian Wang^1,2, Baogui Zhang^1,2,3, Huilou Liang^1,2, Yue Wu^1,2, Jing An⁴, Yan Zhuo^1,2,5, Rong Xue^1,2,6, and Fangrong Zong^7
1State Key Laboratory of Brain and Cognitive Sciences, Beijing MRI Center for Brain Research, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China, ²University of Chinese Academy of Sciences, Beijing, China, ³Brainnetome Center, Institute of Automation, Chinese Academy of Sciences, Beijing, China, ⁴Siemens Shenzhen Magnetic Resonance Ltd., Shenzhen, China, ⁵CAS Center for Excellence in Brain Science and Intelligence Technology, Chinese Academy of Sciences, Beijing, China, ⁶Beijing Institute for Brain Disorders, Beijing, China, ⁷Institute of Biophysics, Chinese Academy of Sciences, Beijing, China

Multidimensional MRI experiments have shown unique significance in resolving the correlations between physical parameters and relaxation properties. This study showed T₁ Relaxation-Diffusion correlation imaging spectra of an in vivo human brain and a multicomponent phantom with optimized acquisitions.  The acquisition paradigm was obtained from genetic algorithm (GA) by maximizing the singular values of the corresponding kernel functions. We detected significant differences in these spectra inversed from different components, which might help to identify tissue microstructure at the sub-voxel resolution. Our study offered a high probability of applying multidimensional MRI techniques in diagnosing neurological diseases.

Jinmin Xu¹, Huafeng Liu¹, and HuiHui Ye^1
1State key Laboratory of Modern Optical Science and Engineering, Zhejiang University, Hangzhou, China

A novel water-fat separated MRF approach was proposed to minimize the known biases introduced by the B1+, B0 and water and fat partial volume. By incorporating the water-fat separation approach in the framework, the dictionary size could be greatly reduced with the known B0 and FF map. Multiple maps of parameters including B1+, B0, FF and T1 and T2* of water and fat can be acquired within 14s for one slice.

Nagomi Fukuda¹, Yuki Kanazawa², Hiroaki Hayashi³, Yuki Matsumoto², Masafumi Harada², Motoharu Sasaki², and Akihiro Haga^2
1Institute of Biomedical Sciences, Tokushima University Graduate School, Tokushima, Japan, ²Graduate school of Health Science, Tokushima University, Tokushima, Japan, ³Faculty of Health Sciences, Institute of Medical, Pharmaceutical and Health Sciences, Kanazawa University, Kanazawa, Japan

We determine the accuracy of our B₁ correction method using a cylindrical digital phantom and a digital brain phantom. Comparing the B₁ correction accuracy of each sample, the relative uncertainties of each T₁ value were less than 5%, and when long TR, the accuracy of B₁ correction was almost equal to short TR. In addition, our method enabled us to correct the slice direction accurately. The proposed method can correct B₁ inhomogeneity with high accuracy to in-plane and slice direction without additional scanning

Naohiro Eda¹, Motofumi Fushimi¹, and Takaaki Nara^1
1The University of Tokyo, Tokyo, Japan

We previously reported a method for magnetic resonance electrical properties tomography (MREPT) based on Helmholtz decomposition at ISMRM 2021. However, the method causes a reconstruction error due to the assumption that the unmeasurable components of the magnetic field are zero. This paper extends our previous method by approximating the unmeasurable components using a homogeneous simulation model that has the same geometry as the object to be examined and has homogeneous electrical properties (EPs). The efficacy of the proposed method was validated through numerical simulations.