Xiwei Li¹, Anliang Chen¹, Ailian Liu¹, and Zhigang Wu^2
1the First Affiliated Hospital of Dalian Medical University, Dalian, China, ²Philips Healthcare, Beijing, China

        Rectal cancer is a malignant tumor with extremely high morbidity and mortality. It will change the physical microenvironments, but it is still a challenge to differentiate the rectal cancer with and without vascular invasion. T2 mapping imaging is a quantitative biomarker with good repeatability and stability. T2 mapping can non-invasively visualize and quantify tissue components (such as edema, fibrosis etc )without contrast agent. This study aims to assess the performance of T2 mapping on differentiating rectal cancer with and without vascular invasion, which may yield higher diagnostic confidence.

Yang Fan¹, Yang Jieyin¹, Sun Jiayu¹, Zhang Xiaoyong², and Ling Chuntang ^3
1Department of Radiology, West China hospital of Sichuan University, Chengdu, China, ²Department of Clinical Science, Philips Healthcare, Chengdu, China, ³Department of Application, Philips Healthcare, Chongqing, China

2D T2-weighted (T2WI) breast MRI is prominently used for the identification of the colliquative necrosis and cysts, and it can also contribute to the characterization of lesions as benign or malignant. However, 2D imaging is routinely used in clinical practice, which has lower resolution, slice gaps, and may suffer distortion to delineate the breast lesions. In this study, we applied the Compressed-sensing Artificial Intelligence (CS-AI) framework to further increase the spatial resolution and reduce the scan time. The results of this study demonstrated that the high-resolution T2-weighted 3D CS-AI can provide further benefits to improve the depiction of diagnostic findings of breast lesions.

Mayumi Takeuchi¹, Kenji Matsuzaki², and Masafumi Harada^1
1Department of Radiology, Tokushima University, Tokushima, Japan, ²Department of Radiological Technology, Tokushima Bunri University, Sanuki-city, Japan

Intra-tumoral hemorrhage is one of the suggestive pathological findings of malignancy. Surgically proven 16 benign and 64 malignant solid or complex ovarian tumors were retrospectively evaluated. High intensity hemorrhagic foci on T1WI were detected in 16 of 64 malignant lesions (25%), whereas signal voids due to hemorrhage on susceptibility-weighted sequence (SWS) were detected in 41 of 64 lesions (64%). Neither high intensity foci on T1WI nor signal voids on SWS was detected in all 16 benign tumors. We conclude that the demonstration of intra-tumoral hemorrhage in patients with ovarian tumors by SWS may provide valuable diagnostic findings.

Chen Lihua¹, Song Qingwei¹, Gao Mingli¹, Wang Jiazheng², Lin Liangjie², Wu Zhigang², and Liu Ailian^1
1The First Affiliated Hospital of Dalian Medical University, DaLian, China, ²Philips Healthcare, Beijing, China

Amide proton transfer-weighted (APTw) imaging is a novel MRI imaging tool for detection of amide protons in mobile cellular proteins and peptides. This study aims to assess and differentiate prostate cancers with and without bone metastasis using APTw MRI and alkaline phosphatase (ALP).Results show that prostate cancers with bone metastasis were associated with significantly higher APTw and ALP values than those without bone metastasis. The diagnostic efficiency is the highest using APT combined with ALP. Therefore, APTw imaging together with ALP may provide an more effective tool to predict the aggressiveness of prostate cancer.

Dan Zhang¹, Zhi Chen¹, Na Song¹, Zhuo Wang¹, Shao Zhang¹, Xiao Chen¹, Xiao Wei², and Bing Chen^1
1The Department of Radiology, The General Hospital of Ningxia Medical University, Yinchuan, China, ²GE Healthcare,MR Research, Beijing, China

In this study, we aim to investigate the diagnostic value of PI-RADS v2.1 and PI-RADS v2 (prostate imaging reporting and data system version 2.1 and version 2) in diagnosing transition zone prostate cancer. It was concluded that the diagnostic value of PI-RADS v2.1was not lower than that of PI-RADS v2.

Dan Zhang¹, Zhi Chen¹, Na Song¹, Zhuo Wang¹, Shao Zhang¹, Xiao Chen¹, Xiao Wei², and Bing Chen^1
1The Department of Radiology, The General Hospital of Ningxia Medical University, Yinchuan, China, ²GE Healthcare,MR Research, Beijing, China

In this study, we aim to investigate the diagnostic value of PI-RADS v2.1 and PI-RADS v2 (prostate imaging reporting and data system version 2.1 and version 2) for peripheral zone prostate cancer. It was concluded that the diagnostic value of PI-RADS v2.1 was higher.

Lili Xu¹, Gumuyang Zhang¹, Li Mao², Xiuli Li², Hao Sun¹, and Zhengyu Jin^1
1Peking Union Medical College Hospital, Beijing, China, ²Deepwise AI Lab, Beijing, China

Accurate prostate segmentation on MR images plays an important role in the management of prostate diseases. Recently proposed deep learning architecture has been successfully applied for medical image segmentation to overcome the shortcomings of manual segmentation. Our study proposed a 3D UNet model for automatic and accurate prostate gland segmentation on both DWI and T2WI images. This model was tested in 3 different external cohorts and showed satisfactory results on T2WI images. The segmentation performance on DWI images was inferior but still inspiring in the external testing group. This study might benefit the management of prostate diseases in the future.

Ashley Wilton Stewart^1,2, Jonathan Goodwin^3,4, Simon Daniel Robinson^2,5,6, Kieran O’Brien^1,2,7, Jin Jin^1,2,7, Markus Barth^1,2,8, and Steffen Bollmann^1,2,8
1ARC Training Centre for Innovation in Biomedical Imaging Technology, The University of Queensland, St Lucia, Australia, ²Centre for Advanced Imaging, The University of Queensland, St Lucia, Australia, ³Department of Radiation Oncology, Calvary Mater Hospital, Newcastle, Australia, ⁴School of Information and Physical Sciences, University of Newcastle, Newcastle, Australia, ⁵High Field MR Center, Department of Biomedical Imaging and Image-Guided Therapy, Medical University of Vienna, Vienna, Austria, ⁶Department of Neurology, Medical University of Graz, Graz, Austria, ⁷Siemens Healthcare Pty Ltd, Brisbane, Australia, ⁸School of Information Technology and Electrical Engineering, The University of Queensland, St Lucia, Australia

Gold fiducial markers (FMs) for prostate radiotherapy are commonly localized using Computed Tomography (CT), though interest in MR-only workflows is growing. One current limitation with magnitude-based MR-only workflows is the ability to distinguish FMs from blood products and calcifications because all appear as signal voids. Quantitative Susceptibility Mapping (QSM) separates sources based on their susceptibility, and may be a potential solution. We apply the QSMxT pipeline to gradient-echo images of a prostate cancer patient with FMs and a calcification. We found that QSM provided a contrast that could differentiate the FMs from calcifications to a degree comparable to CT.

Dan Zhang¹, Zhi Chen¹, Na Song¹, Zhuo Wang¹, Shao Zhang¹, Xiao Chen¹, Xiao Wei², and Bing Chen^1
1The Department of Radiology, The General Hospital of Ningxia Medical University, Yinchuan, China, ²GE Healthcare,MR Research, Beijing, China

In this study, we aim to investigate the value of DISCO quantitative parameters combined with MUSE in distinguishing benign and malignant prostate lesions. It was concluded that Ktrans, Kep, Ve and ADC can be used as imaging biomarkers to distinguish benign and malignant prostate lesions; Ktrans and ADC are helpful to predict aggressiveness of the tumor，and can be used as independent predictors of prostate cancer. 

Wei Wang¹, Jing Liu¹, Wei Li¹, Ke Xue², Yongming Dai², and Jianxing Qiu^1
1Peking University First Hospital, Beijing, China, ²United Imaging Healthcare, Shanghai, China

This study is targeted to compare the performance of three-dimensional (3D) breath-hold gradient and spin-echo sequence (GRASE) and conventional respiratory-trigger fast spin-echo sequence (FSE) for magnetic resonance urography (MRU). Image quality and diagnostic performance of urinary tract dilation from two data sets were evaluated by three radiologists. The results showed that 3D GRASE MRU had a better performance compared with the FSE MRU, indicating 3D GRASE technique could be a feasible option for 3D MRU.

Gumuyang Zhang¹, Li Chen¹, Hailong Zhou¹, Yunna Wang¹, Jinxia Zhu¹, Marcel Dominik Nickel², Elisabeth Weiland³, Hao Sun¹, and Zhengyu Jin^1
1Peking Union Medical College Hospital, Beijing, China, ²Siemens Healthineers Ltd., Beijing, China, ³MR Application Predevelopment, Siemens Healthcare GmbH, Erlangen, Germany

This study evaluated deep learning accelerated T2w imaging (T2_DL) of the bladder in terms of acquisition time (TA), overall image quality, presence of artifacts, diagnostic confidence, sharpness of lesions, and VI-RADS T2 score compared to a standard T2w (T2_S) sequence in twenty-five patients. Two radiologists evaluated the images independently. TA of T2_DL was reduced by nearly 50% compared to T2_S. Overall image quality and sharpness of lesions were superior in T2_DL, while artifacts, diagnostic confidence and T2 score were similar between T2_S and T2_DL. T2_DL has the potential to replace T2_S in bladder MR.

Ming-Cheng Liu^1,2, Yi-Jui Liu^2,3, Pin-Sian Lyu³, Guan-Xin Pan³, Si-Wa Chan^4,5,6, Yen-Ting Lin^1,7, Siu-Wan Hung^1,8,9, Jyh-Wen Chai¹, and Kun-Yuan Chiu^10,11
1Radiology, Department of Radiology, Taichung Veterans General Hospital, Taiwan, Taichung, Taiwan, ²Ph.D. Program of Electrical and Communications Engineering, Feng Chia University, Taichung, Taiwan, Taichung, Taiwan, Taiwan, ³Department of Automatic Control Engineering, Feng Chia University, Taichung, Taiwan, Taichung, Taiwan, ⁴Department of Medical Imaging, Taichung Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Taichung, Taiwan, Taichung, Taiwan, ⁵School of Medicine, Tzu Chi University, Hualien, Taiwan, Taichung, Taiwan, ⁶Department of Medical Imaging Radiological Science, Central Taiwan University of science and Technology, Taichung, Taiwan, Taichung, Taiwan, ⁷Institute of Clinical Medicine, National Yang-Ming University, Taiwan, Taichung, Taiwan, ⁸School of Medical Imaging and Radiological Sciences, Chung Shan Medical University, Taiwan, Taichung, Taiwan, ⁹Department of Veterinary Medicine, National Chung Hsing University, Taiwan, Taichung, Taiwan, ¹⁰Division of Urology, Department of Surgery, Taichung Veterans General Hospital, Taiwan, Taichung, Taiwan, ¹¹Department of Applied Chemistry, National Chi Nan University, Taiwan, Taichung, Taiwan

Tumor necrosis, hemorrhage, and cystic change may influence the result of MR histogram analysis between ccRCC and ncRCC. We conducted EP ROI and WL ROI measurements of these two tumors. Five most important metrics were found in differentiation, including three EP ROI metrics on the T1 A and V phase, and the T2W images, and two WL ROI metrics on the DWI b800 images and the T2W images. Both EP ROI and WL ROI played a role in the diagnosis ccRCC and ncRCC. The addition of these two ROI measurements of histogram analysis to routine MRI evaluation may benefit diagnosis.

You-Zhen Feng¹, Zhong-Yuan Cheng¹, Long Qian², and Xiang-Ran Cai^1
1Medical Imaging Center, the First Affiliated Hospital of Jinan University, Guangzhou, China, ²MR Research,GE Healthcare, Beijing, China

Apply DKI technology to the assessment of early renal function changes in diabetic patients. These results indicate that the microstructure of kidney tissue may have changed in diabetic patients without clinically impaired renal function. In addition, the results of correlation analysis between DKI parameters and clinical laboratory indicators of renal function (HBA1, Uscr, and eGFR) show that there is a linear correlation between most of the parameters, which objectively confirms the feasibility of DKI parameters to reflect changes in renal function.

Liang Pan¹, Jie Chen¹, Wei Xing¹, and Jilei Zhang^2
1Third Affiliated Hospital of Soochow University, Changzhou, China, ²Philips Healthcare, Shanghai, China

Transplant renal artery stenosis (TRAS) is a less common but important etiology of renal dysfunction following transplantation. It is of great importance to include TRAS in the differential diagnosis for allograft dysfunction early after transplantation. We used ASL combined with 2D-PC Q-flow MRI to assess hemodynamic changes of TRAS. The results reflected that there were differences in the mean flux and velocity of the stenotic segment of transplant renal artery and renal allograft perfusion between different degrees of TRAS, and the mean flux and velocity were more closely related to the degree of TRAS compared with renal allograft perfusion.

qiqi zhou¹, Weiyin Vivian Liu², Qian Tang³, Ling Song¹, Chao Liu¹, Hu Chen¹, and Lin Xu^1
1Department of Radiology,Taihe hospital, Hubei  University  of  Medicine, Hubei, China, ²GE Healthcare, Beijing, China, ³Biomedical Engineering College, Hubei University of Medicine, Hubei, China

It is of importance to acquire diffusion-weighted images with high image quality in a short time whereas higher b value and a greater number of b values reduce image quality and prolong scan time. In this study, we aimed to build an optimized protocol for IVIM in diagnosis of prostate cancer.We observed mean of all ADC, D, D* and f derived from IVIM with 5 b-value scheme showed almost the same diagnosis efficiency to IVIM with 13 b-value scheme, suggesting sensitivity and specificity were not different between different b-value scheme.