Jia Liu¹, Yukun Cao¹, Kuikui Zhu¹, Sheng Yao², Mei Yuan¹, Xiangchuang Kong¹, Xiaoming Liu¹, Yumin Li¹, Yue Cui¹, Xiaoyu Han¹, Xiaoyue Zhou³, Rui Meng¹, and Heshui Shi^1
1wuhan union hospital, wuhan, China, ²wuhan tongji hospital, wuhan, China, ³Siemens Healthcare, shanghai, China

In our study, we conducted a prospective study to investigate the subclinical cardiotoxicity by serial cardiovascular magnetic resonance (CMR) in patients receiving immune checkpoint inhibitor (ICI). Some functional parameters and global strains of bi-ventricles revealed significant decrease at the early stage on ICI therapy, and ejection fraction as well as global strain values of both ventricles all showed a decline trend. Tissue characterization is an extra advantage of CMR technique. While in our study, parameters (T1 mapping, T2 mapping and late gadolinium enhancement) regarding tissue characterization of global myocardium showed no significant differences, compared with baseline data.

Shiqin Yu¹, Xiuyu Chen¹, Kai Yang¹, Jiaxin Wang¹, Kankan Zhao², Wenhao Dong¹, Weipeng Yan¹, and Shihua Zhao^1
1Fuwai Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China, ²Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, SZ University Town, Shenzhen, China

In patients with prominent excessive trabeculation, left ventricular systolic dysfunction was detected early by cardiac MRI feature tracking despite the presence of normal left ventricular ejection fraction and was associated with excessive trabecular complexity assessed by fractal dimension.

Yinghao Zhang¹, Yue Hu¹, and Xin Lu^2
1School of Electronics and Information Engineering, Harbin Institute of Technology, Harbin, China, ²School of Computer Science and Informatics, De Montfort University, Leicester, United Kingdom

Low-rank tensor models have been applied in accelerating dynamic magnetic resonance imaging (dMRI). Recently, a new tensor nuclear norm based on t-SVD has been proposed and applied to tensor completion. Inspired by the different properties of the tensor nuclear norm (TNN) and the Casorati matrix nuclear norm (MNN), we introduce a novel dMRI reconstruction method combining TNN and Casorati MNN, which we term as TMNN. Moreover, we convert the the TMNN dMRI reconstruction problem into a simple tensor completion problem, which can be efficiently solved by the alternating direction method of multipliers (ADMM).

Mei Yuan¹, Jia Liu¹, Huan Liu², Xiaoyue Zhou³, and Heshui Shi^1
1Department of Radiology, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China, ²GE Healthcare, Shanghai, China, ³Siemens Healthineers Ltd., Shanghai, China

本研究探索了原生T1映射的纹理分析（TA），以便检测非apic肥大性心肌病（HCM）、心肌梗塞和高血压心脏病（HHD）的特定模式，具有良好的诊断效果，可反映心肌纤维化。结果表明，TA能够有效地区分和显示与心脏功能相媲美的诊断精度，并结合菌株分析。TA、心脏功能和应变分析的结合，使非 apical HCM 或 apical HCM 与 HHD 的区分具有最佳诊断精度。这表明，将 TA 与心脏功能和应变参数相结合，比单独使用 TA 具有增量值。

Baiyan Zhuang¹, Minjie Lu¹, and Shihua Zhao^1
1Fuwai Hospital, State Key Laboratory of Cardiovascular Disease，PUMC, Beijing, China

Given the fact that patients who have underwent repair of tetralogy of fallot (rTOF) further need pulmonary valve replacement (PVR) due to pulmonary regurgitation, but the effect of PVR is mixed.  We prospectively enrolled 45 asymptomatic rTOF patients who required PVR due to moderate or severe pulmonary regurgitation and measured their pre-operative strain and strain rate. After follow-up and analysis, we found that the radial strain (RS), circumferential strain (CS), longitudinal strain (LS), systolic radial strain rate (RSRs), early diastolic radial strain rate (RSRe) and early diastolic longitudinal strain rate (LSRe) before PVR are important prognostic factors for adverse events.

Yasuhiro Goto¹, Nagao Michinobu², Masami Yoneyama³, Johannes M Peeters⁴, Isao Shiina¹, Kazuo Kodaira¹, Yutaka Hamatani¹, Takumi Ogawa¹, Mana Kato¹, and Shuji Sakai^2
1Department of Radiological Services, Tokyo Woman's Medical University, Tokyo, Japan, ²Department of Diagnostic imaging & Nuclear Medicine, Tokyo Woman's Medical University, Tokyo, Japan, ³Philips Japan, Tokyo, Japan, ⁴Philips Healthcare, Best, Netherlands

We evaluated the feasibility of the combined use of MoSE CINE imaging (to determine exact TD as a preparation procedure), MoCo-TSE-DWI acquisition and FEIR post-processing (to register the respective diffusion images) for improving the robustness of quantitative cardiac IVIM mapping. Combined use of the MoSe-CINE procedure for determining the exact cardiac trigger delay, MoCo-TSE-DWI acquisition, and FEIR post-processing to register among all source images, could be useful to increase the robustness of image quality and improve the quantitative accuracy in myocardial IVIM mapping.

xin Zhang¹, Lianggeng Gong¹, jingyang Wen¹, long Qian², and Weiyin Vivian Liu^2
1The Second Affiliated Hospital of Nanchang University, Nanchang, China, ²MR Research, GE Healthcare, Beijing, China

Pathological changes vary in the morphology and function during the process of hypertrophic cardiomyopathy. In this study, we compared the diagnostic value of  fractal analysis and tissue tracking on cardiovascular magnetic resonance images to characterize the trabecular complexity as myocardial strain alteration for HCM.

Yujie Gao¹, Xiaoyue Zhou², Shengen Liao³, and Yi Xu^1
1Department of Radiology, The First Affiliated Hospital of Nanjing Medical University, Nanjing, China, ²MR Collaboration, Siemens Healthcare, Shanghai, China, ³Department of Cardiology, The First Affiliated Hospital of Nanjing Medical University, Nanjing, China

This study investigated the relationship between cardiac magnetic resonance (CMR) T1 mapping parameters and N-terminal pro-B-type natriuretic peptide (NT-proBNP) in non-ischaemic dilated cardiomyopathy (DCM) patients. The results showed that, in DCM, a non-linear relationship existed between NT-proBNP and septal native T1, indicating a complex relationship between the secretion of NT-proBNP and myocardial diffuse fibrosis. When NT-proBNP is greater than the inflection point(1105 pg/ml), NT-proBNP may provide additional value for the judgment of myocardial diffuse fibrosis.

Gang Zheng¹, Yifang Li², Ekaterina Salimova ¹, Hong X Wang¹, Shenjun Zhong ¹, Chrishan S Samuel ², Michael de Veer¹, and Gary Egan ^1
1Monash Biomedical Imaging, Monash University, Clayton, Australia, ²Monash Biomedicine Discovery Institute and Department of Pharmacology, Monash University, Clayton, Australia

The detection and localization of diffuse myocardial fibrosis is for the non-invasive evaluation of therapeutic strategies in murine cardiac disease models. Cardiac late-gadolinium enhancement (LGE) and T1 mapping are two techniques for non-invasive cardiovascular magnetic resonance (CMR) imaging. This abstract demonstrates that the magnetization prepared 2 rapid acquisition gradient echo (MP2RAGE) can simultaneously acquire LGE and T1 maps for the detection of diffuse myocardial fibrosis in a 1K/DOCA/salt-induced murine model of hypertensive disease.

Lingping Ran¹, Lu Huang¹, Xianghu Yan¹, Yi Luo¹, Shuheng Zhang², Shiyu Zhang², Yuan Zheng³, Jian Xu³, and Liming Xia^1
1Department of Radiology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China, ²United Imaging Healthcare, Shanghai, China, ³UIH America, Inc., Houston, TX, United States

High resolution T2-weighted dark blood (HR-T2W-DB) imaging is not always robust for clinical use because of low SNR and long scan time. The purpose of this study was to evaluate a novel deep learning (DL) based reconstruction method in T2W-DB sequence that achieves higher spatial resolution and same scan duration compared with traditional reconstruction method. Quantitative and qualitative image assessment demonstrated that DL based HR-T2W-DB sequence showed better CNR in region of edema and LV free wall visibility, which might help detecting myocardial edema.

Binghua Chen¹, Jilei Zhang², Weibo Chen², Yihong Zhang², Jianrong Xu¹, and Lianming Wu^1
1Renji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China, ²Philips Healthcare, Shanghai, China

We propose an automatic diagnostic model for left ventricular aneurysm after myocardial infarction. A epicardium segmentation model was established by mixing 2- and 3-chamber images of 90 healthy volunteers, and the dice in all cohorts exceeded 0.95. Five heartphaseimages around end-systolic and end-diastolic stages, multiplied by the predicted mask using the segmentation model were used as the input of the classification model. Data from 259 AMI patients were divided into training cohort (206) and independent testing cohort (53). ResNet was selected to extract the features of 2- and 3- chamber data. Finally, AUC achieved on 0.987/0.946 in training and testing cohort. The automtic deep learning model can be used as a scheme to dignose LVA after AMI, and has potential clinical value for early detection of the risk of the ventricular aneurysm in AMI patients.

Yujiao Deng^1,2 and Jing Chen^2
1Department of Nuclear Medicine, Sichuan Academy of Medical Sciences & Sichuan Provincial People’s Hospital, chengdu,sichuan, China, ²Department of Radiology, The Affiliated Hospital of Southwest Medical University, Luzhou, Sichuan, China

Regular exercise can have a beneficial effect on the cardiovascular system, but more and more reports of sudden cardiac deaths in athletes and general people during exercise had made people gradually pay attention to the damage effect of exercise on the heart.There is currently lack of imaging examination methods for early and accurate judgment of athletes’ myocardial injury.This study intended to use intravoxel incoherent motion diffusion-weighted imaging to evaluate the myocardial microcirculation of athletes to early predict the incidence of adverse cardiac events such as cardiac remodeling and delayed enhancement of the myocardium.

Fang Wang¹, Zhen Zhang¹, Xiuzheng Yue², Xiaowei Ruan¹, Rongrong Zhu¹, Ruoshui Ha¹, and Yanbin Yang^1
1Medical Imaging Center, People's Hospital of Ningxia Hui Autonomous Region, Yinchuan, China, ²Philips Healthcare, Beijing, China

The study is to determine clinical utility of T1rho to detect patients with cardiomyopathy. In the LGE positive group, the T1 mapping values and ECV values in the lesion area were significantly higher than those in the LGE negative group and the normal control group, while the T1rho values in the lesion area were significantly lower than those in the LGE negative and normal groups, and the display range of lesions was in good consistency with LGE and ECV. It can be used as an endogenous diagnostic index when myocardial structure changes.

Yanan Zhao¹, Tao Li¹, Jianing Cui¹, Xueqian Liu², and Xiuzheng Yue^3
1Department of Radiology, Chinese People’s Liberation Army General Hospital, Beijing, China, ²Qinhuangdao Workers' Hospital, Qinhuangdao, China, ³Philips Healthcare, Beijing, China

The reduced right ventricular ejection fraction (RVEF) has prognostic value for predicting cardiovascular mortality in patients with left ventricular acute myocardial infarction (LVAMI). Previous studies have shown that interaction between left ventricular (LV) and right ventricular (RV). This study used CMR to further explore what factors affect the RVEF. The results shown that the incidence of reduced RVEF was 40% in LVAMI patients and the occurrence of reduced RVEF was associated with LVEF, LV end-systolic volume index and RV end-systolic volume index.

Takashige Yoshida¹, Kohei Yuda², Masami Yoneyama³, Jihun Kwon³, and Marc Van Cauteren^4
1radiology, Tokyo metropolitan police hospital, Tokyo, Japan, ²Tokyo metropolitan police hospital, Tokyo, Japan, ³Philips Japan, Tokyo, Japan, ⁴Philips Healthcare, Best, Netherlands

High-resolution imaging and T1 mapping is needed to achieve useful clinical information optimally in cardiac MRI. However, prolonged acquisition time can lead to poor or non-diagnostic image quality. In this study, we investigated the use of a deep learning-based reconstruction algorithm to highly accelerate T1map acquisition for cardiac MRI. Adaptive-CS-Net, a deep neural network previously introduced at the 2019 fastMRI challenge, was expanded and integrated into the Compressed-SENSE Artificial Intelligence (CS-AI) reconstruction. The purpose of this study was to compare the image quality of high-resolution T1map between reference and accelerated methods: SENSE, Compressed-SENSE, and CS-AI.