Peter Wei¹, Anugayathri Jawahar¹, Daniel V. Litwiller², and Andreas M. Loening^1
1Department of Radiology, Stanford University, Stanford, CA, United States, ²Global MR applications and Workflow, GE Healthcare, New York, NY, United States

A steady-state free precession (SSFP) sequence is used in many MR enterography (MRE) protocols for acquiring cine images to assess bowel motility, inflammation, and strictures. However, SSFP suffers from susceptibility and banding artifacts that become more significant at high field strengths. In this IRB approved retrospective study, we compared a cine SSFP sequence to a cine T2-weighted single-shot fast spin echo (SSFSE) sequence in 41 patients. We found SSFSE demonstrated significantly superior subjective assessments of image quality, improved diagnostic performance compared to cine SSFP, and successfully mitigated SSFP artifacts that can otherwise limit the exam.

Guangtao Chen¹, Hing-Chiu Chang¹, and Keith Wan-Hang Chiu^1
1The Department of Diagnostic Radiology, The University of Hong Kong, Hong Kong, China

Diffusion-weighted imaging(DWI) has been shown to be useful in evaluation of active bowel inflammation. Single-shot diffusion-weighted echo-planar imaging (ssDW-EPI) was commonly used in previous studies. However, ssDW-EPI suffers from geometric distortion and low spatial resolution. Moreover, the data acquisition of ssDW-EPI in bowel region become more difficult due to serve off-resonance effect. A recently developed multiplexed sensitivity-encoding (MUSE) framework can produce multi-shot DW-EPI (msDW-EPI) data with improved spatial resolution and reduced distortions. In this study, we demonstrated that the higher resolution and better overall image quality of msDW-EPI using MUSE framework potentially increased the accuracy in diagnosing active bowel inflammation.

Vinay Prabhu¹, Nicolas Gillingham¹, Mary T. Bruno¹, James Babb¹, Henry Rusinek¹, and Hersh Chandarana^1
1Radiology, NYU Langone Health, New York, NY, United States

We studied the robustness of liver MRI texture features by scanning five healthy volunteers at 3T, first using standard institutional acquisition parameters, and then introducing slight variation in acquisition parameters. Our results demonstrate that a number of texture features were not robust to acquisition parameter changes.

Shinya Kojima^1,2, Daiki Tamada ², Tetsuya Wakayama ³, Shintaro Ichikawa ², Hiroyuki Morisaka ⁴, Shigeru Suzuki ¹, and Utaroh Motosugi ^2
1Department of Radiology, Tokyo Women’s Medical University Medical Center East, Arakawa, Japan, ²Department of Radiology, University of Yamanashi, Yamanashi, Japan, ³MR Collaboration and Development, GE Healthcare, Hino, Japan, ⁴Department of Radiology, Saitama Medical University International Medical Center, Saitama, Japan

Motion artifact by irregular respiration disturbs accurate diagnosis in dynamic contrast-enhanced MRI of the liver. We developed a motion artifact reduction algorithm using a convolutional network (MARC). The training was performed using U-net with the arterial phase images with and without simulated artifacts. For verifying the ability of MARC algorithm, contrast-to-noise ratio measurement and visual assessment were performed in 120 cases. The image quality of arterial phase images with motion artifacts were significantly improved after applying MARC algorithm, while no particular difference was observed in the images without motion artifacts. MARC provides motion artifacts reduction without variation of image contrast.

Xinzeng Wang¹, Daniel Litwiller², Marc Lebel³, Ali Ersoz⁴, Lloyd Estkowski⁴, Jason Stafford⁵, and Ersin Bayram^6
1GE Healthcare, Houston, TX, United States, ²Global MR Applications & Workflow, GE Healthcare, New York, NY, United States, ³Global MR Applications & Workflow, GE Healthcare, Calgary, AB, Canada, ⁴Global MR Applications & Workflow, GE Healthcare, Waukesha, WI, United States, ⁵Department of Imaging Physics, MD Anderson Cancer Center, Houston, TX, United States, ⁶Global MR Applications & Workflow, GE Healthcare, Houston, TX, United States

T2W FSE-PROPELLER is robust to susceptibility artifacts and bulk motion, but requires longer acquisition times compared to conventional FSE methods. Recently, a reduced Field-Of-View PROPELLER sequence using rotating outer volume suppression method has been proposed and optimized to reduce the scan time for small FOV and high-resolution T2W imaging. However, image SNR is comparatively lower compared to the conventional PROPELLER with phase oversampling. In this work, a deep learning based PROPELLER reconstruction method was used to improve the SNR and image quality of the reduced Field-Of-View PROPELLER.

David Hoar¹, Peter Lee², Alessandro Guida³, Steven Patterson³, Chris Bowen^3,4, Jennifer Merrimen⁵, Cheng Wang⁵, Ricardo Rendon⁶, Steven Beyea^3,4, and Sharon Elizabeth Clarke^3,4
1Department of Electrical and Computer Engineering, Dalhousie University, Halifax, NS, Canada, ²Faculty of Computer Science, Dalhousie University, Halifax, NS, Canada, ³Biomedical Translational Imaging Centre, Nova Scotia Health Authority and IWK Health Centre, Halifax, NS, Canada, ⁴Diagnostic Radiology, Dalhousie University, Halifax, NS, Canada, ⁵Pathology, Dalhousie University, Halifax, NS, Canada, ⁶Urology, Dalhousie University, Halifax, NS, Canada

The purpose of this study was to develop a convolutional neural network (CNN) for dense prediction of prostate cancer using mp-MRI datasets. Baseline CNN outperformed logistic regression and random forest models. Transfer learning and unsupervised pre-training did not significantly improve CNN performance; however, test-time augmentation resulted in significantly higher F1 scores over both baseline CNN and CNN plus either of transfer learning or unsupervised pre-training. The best performing model was CNN with transfer learning and test-time augmentation (F1 score of 0.59, AUPRC of 0.61 and AUROC of 0.93).

Shi yun SUN¹, Zhuo lin Li¹, Ying ying Ding¹, Yi fan Liu¹, Dong xue ZHANG¹, Li sha NIE², Ke XUE¹, and Dian Ke DU^1
1Radiology, Yunnan Cancer Hospital，The Third Affiliated Hospital of Kunming Medical University, Kunming, China, ²GE Healthcare, MR Research China, Beijing, China, China

It is reported that dynamic contrast imaging and T2 relaxation time can be used to differentiate benign and malignant breast lesions. However, few researches have investigated T1 and T2 relaxation time changes before and after contrast injection. But it's important for the diagnosis of breast diseases. Thus, the study aims to utilize the T1 and T2mapping in synthetic MR to differentiate benign and malignant lesions. Our results demonstrated that T1 and T2 mapping could constitute a new adjunct in the MRI diagnosis of breast diseases.

Jong Bum Son¹, Ken-Pin Hwang¹, Marion E. Scoggins², Basak E. Dogan³, Gaiane M. Rauch², Mark D. Pagel⁴, and Jingfei Ma^1
1Imaging Physics Department, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, ²Diagnostic Radiology Department, The University of Texas MD Anderson Cancer Center, Houston, TX, United States, ³Department of Diagnostic Radiology, The University of Texas Southwestern Medical Center, Dallas, TX, United States, ⁴Cancer Systems Imaging Department, The University of Texas MD Anderson Cancer Center, Houston, TX, United States

A Dixon conditional generative adversarial network (DixonCGAN) was developed for Dixon water and fat separation. For the robust water image reconstruction, DixonCGAN performs water and fat separation with three processing steps: (1) phase-correction with DixonCGAN, (2) error-correction for DixonCGAN processing, and (3) the final water and fat separation. A conditional generative adversarial network (CGAN) originally designed to change photo styles could be successfully modified to perform phase-correction with improved global and local image details. Moreover, localized deep-learning processing errors could be effectively recovered with the proposed deep-learning error-correction processes.

Catherine Judith Moran¹, Christopher M Sandino², Joseph Cheng¹, Marcus T Alley¹, Bruce Daniel¹, and Brian A. Hargreaves^1
1Radiology, Stanford University, Stanford, CA, United States, ²Electrical Engineering, Stanford University, Stanford, CA, United States

Breast MRI without a contrast injection has the potential to increase accessibility and compliance to the method in women who are recommended to undergo annual MRI to screen for breast cancer. Steady-state diffusion weighted methods provide robust image quality in comparison to conventional diffusion weighted methods which can suffer from variable image quality due to distortion, blurring and low-resolution. Double Echo Steady State (DESS) acquisition with a cones k-space trajectory is investigated for non-contrast enhanced breast MRI in 30 women undergoing clinically indicated breast MRIs.
 

Mami Iima^1,2, Masako Kataoka¹, Maya Honda¹, Ayami Ohno Kishimoto¹, Rie Ota¹, Akane Ohashi¹, Yuta Urushibata³, Thorsten Feiweier⁴, Masakazu Toi⁵, and Kaori Togashi^1
1Diagnostic Imaging and Nuclear Medicine, Kyoto University Graduate School of Medicine, Kyoto, Japan, ²Clinical Innovative Medicine, Institute for Advancement of Clinical and Translational Science, Kyoto University Hospital, Kyoto, Japan, ³Siemens Healthcare K.K., Tokyo, Japan, ⁴Siemens Healthcare GMBH, Erlangen, Germany, ⁵Breast Surgery, Kyoto University Graduate School of Medicine, Kyoto, Japan

We investigated the variation of ADC values obtained at diffusion times that are clinically available for differentiation of human breast tumors. The ADC values in both malignant and benign breast tumors decreased with increased diffusion time, and a larger change in ADC values was found in malignant tumors. The significant association found between ADC change and Ki-67 expression might indicate the potential of diffusion time-dependent ADC values as a tool to differentiate these prognostic biomarkers and assess tumor heterogeneity without the need for contrast agents.

Nada Mufti^1,2,3, Michael Aertsen⁴, Michael Ebner^2,3, Lucas Fidon², Tom Vercauteren^2,5, Luc De Catte⁵, Philippe Demaerel⁴, Jan Deprest^1,5, Sebastien Ourselin², Anna L David^1,6, and Andrew Melbourne^2,3
1Institute for Women's Health and Department of Medical Physics and Biomedical Engineering, University College London (UCL), London, United Kingdom, ²School of Biomedical Engineering and Imaging Sciences (BMEIS), King's College London, London, United Kingdom, ³Department of Medical Physics and Biomedical Engineering, University College London (UCL), London, United Kingdom, ⁴Department of Radiology, University Hospitals Katholieke Universiteit (KU) Leuven, Leuven, Belgium, ⁵Department of Obstetrics and Gynaecology, University Hospitals Katholieke Universiteit (KU) Leuven, Leuven, Belgium, ⁶University Hospitals KU Leuven, Leuven, Belgium

Comprehensive evaluation of the fetal central nervous system (CNS) is required to select the most suitable candidates, to counsel parents about fetal spina bifida surgery, and to monitor post-op response. In children and adolescents with spina bifida, MRI assessment of gyrification correlates with motor and cognitive function. Our aim is to determine if MRI can quantify fetal brain gyrification and folding before and after fetal spina bifida surgery. If successful, mapping the gyrification changes to different lobes in the brain may prove useful in the prediction of motor and cognitive function after fetal surgery.

Ryan J Pewowaruk¹, David R Rutkowski², Cody J Johnson², Colin K Kim², Shane A Wells², Wade A Bushman³, Diego Hernando^2,4, and Alejandro Roldán-Alzate^1,2,5
1Biomedical Engineering, University of Wisconsin, Madison, WI, United States, ²Radiology, University of Wisconsin, Madison, WI, United States, ³Urology, University of Wisconsin, Madison, WI, United States, ⁴Medical Physics, University of Wisconsin, Madison, WI, United States, ⁵Mechanical Engineering, University of Wisconsin, Madison, WI, United States

Male urogenital disease is a common problem, and non-invasive methods for diagnosis and disease progression tracking are limited. This study was aimed at developing an MRI-based method to characterize urogenital tissue morphology, bladder-prostate interaction, and blood flow perfusion.  These methods were tested in three patients with BPH and three healthy volunteers. Strong correlation between void fraction and prostate volume was found. Future work will be aimed at applying these methods to larger cohorts so that clinical utility may be further understood.

Yoshiharu Ohno^1,2, Masao Yui³, Daisuke Takenaka⁴, Yoshimori Kassai³, Kazuhiro Murayama¹, and Takeshi Yoshikawa^2
1Radiology, Fujita Health University School of Medicine, Toyoake, Japan, ²Radiology, Kobe University Graduate School of Medicine, Kobe, Japan, ³Canon Medical Systems Corporation, Otawara, Japan, ⁴Diagnostic Radiology, Hyogo Cancer Center, Akashi, Japan

    No one directly compare this new technique with quantitatively assessed CT for pulmonary functional loss evaluation and The Global Initiative for Chronic Obstructive Lung Disease (GOLD) classification in smokers.  We hypothesized that regional ΔT1 change from 3D O₂-enhanced MRI has a potential for pulmonary functional loss assessment and clinical stage classification as well as quantitatively assessed thin-section CT in smokers.  The purpose of this study was to prospectively and directly compare the quantitative capability for pulmonary functional loss assessment and clinical stage classification between 3D O₂-enhanced MRI and thin-section CT in smokers.