Qinwei Zhang¹, Bram F. Coolen¹, Gustav J. Strijkers², Laurens van Buuren³, Uulke van der Heide³, Oliver J. Gurney-Champion ¹, Sónia I. Gonçalves⁴, and Aart J. Nederveen^1
1Department of Radiology, Academic Medical Center, University of Amsterdam, Amsterdam, Netherlands, ²Biomedical Engineering and Physics, Academic Medical Center, University of Amsterdam, Amsterdam, Netherlands, ³Department of Radiation Oncology, The Netherlands Cancer Institute, Amsterdam, Netherlands, ⁴Institute for Biomedical Imaging and Life Sciences, University of Coimbra, Coimbra, Portugal

Diffusion imaging is part of the standard MR imaging protocol for prostate cancer diagnosis. Conventional echo planar imaging (EPI) diffusion sequence has limitation on image resolution and additionally suffers from image distortion. The present study introduces a new stimulated echo based 3D diffusion preparation turbo spin echo sequence (DPsti-TSE) to achieve high-resolution and distortion free image. The sequence is also proved to be immune to eddy currents. 

Zheng Han¹, Yajuan Li¹, and Zheng-Rong Lu^1
1Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States

Prostate cancer (PCa) is the second most lethal cancer in American men with a high incidence rate. Current method of PCa screening is not specific to aggressive cancer type, which results in overtreatment with serious adverse effects. We developed a MRI contrast agent, ZD2-Gd(HP-DO3A), that targets to overexpressed extradomain-B in aggressive PCa. Our result showed an increased sensitivity for MRI detection of aggressive PCa using ZD2-Gd(HP-DO3A), compared with the clinical control agent ProHance®.  This contrast agent can potentially facilitate accurate risk stratification and clinical management of PCa. 

Edward William Johnston¹, Eleftheria Panagiotaki², Elisenda Bonet-Carne², Nicola Stevens¹, David Atkinson¹, Daniel Alexander², and Shonit Punwani^1
1UCL Centre for Medical Imaging, London, United Kingdom, ²UCL Centre for Medical Image Computing, London, United Kingdom

VERDICT (Vascular, Extracellular, and Restricted Diffusion for Cytometry in Tumours) is a microstructural imaging technique that has shown significant potential in preclinical and pilot studies. However, its technical repeatability is unknown and must be established for translational and clinical application.  

 

5 patients underwent consecutive VERDICT acquisitions, and their quantitative parametric maps were compared in tumour and non-tumour regions. We found that cellularity was the most reliable parameter, with almost perfect repeatability in both normal and cancerous prostate tissue. Intra and extracellular volume fractions also performed well, with almost perfect repeatability in the normal prostate and excellent repeatability in cancerous tissue.

Wei Huang¹, Yiyi Chen¹, Andriy Fedorov², Xia Li³, Guido Jajamovich⁴, Dariya I Malyarenko⁵, Madhava Aryal⁵, Peter S LaViolette⁶, Matthew J Oborski⁷, Finbarr O’Sullivan⁸, Richard G Abramson⁹, Mark Muzi¹⁰, Kourosh Jafari-Khouzani ¹¹, Aneela Afzal¹, Alina Tudorica¹, Brendan Moloney¹, Cecilia Besa⁴, Jayashree Kalpathy-Cramer¹¹, James M Mountz⁷, Charles M Laymon⁷, Kathleen Schmainda⁶, Yue Cao⁵, Thomas L Chenevert⁵, Bachir Taouli⁴, Thomas E Yankeelov⁹, Fiona Fennessy², and Xin Li^1
1Oregon Health & Science University, Portland, OR, United States, ²Brigham and Women’s Hospital and Harvard Medical School, Boston, MA, United States, ³General Electric Global Research, Niskayuna, NY, United States, ⁴Icahn School of Medicine at Mount Sinai, New York, NY, United States, ⁵University of Michigan, Ann Arbor, MI, United States, ⁶Medical College of Wisconsin, Milwaukee, WI, United States,⁷University of Pittsburgh, Pittsburg, PA, United States, ⁸University College Cork, Cork, Ireland, ⁹Vanderbilt University, Nashville, TN, United States, ¹⁰University of Washington, Seattle, WA, United States,¹¹Massachusetts General Hospital and Harvard Medical School, Boston, MA, United States

Dynamic Contrast-Enhanced MRI (DCE-MRI) pharmacokinetic modeling is widely used to extract tissue specific quantitative parameters.  However, the accuracy and precision of these parameters can be affected by many factors, with arterial input function (AIF) determination being a primary source of uncertainties. In this multicenter study, we sought to evaluate variations in DCE-MRI parameters estimated from shared prostate DCE-MRI data as a result of differences in AIFs.

Xiaobing Fan¹, Shiyang Wang¹, Milica Medved ¹, Tatjana Antic², Serkan Guneyli ¹, Aytekin Oto ¹, and Gregory S Karczmar ^1
1Radiology, University of Chicago, Chicago, IL, United States, ²Pathology, University of Chicago, Chicago, IL, United States

Accurate measurements of the arterial input function (AIF) are needed in pharmacokinetic models to analyze dynamic contrast enhanced (DCE) MRI data. The AIF often cannot be accurately measured due to T2* and water exchange effects. Therefore, population AIFs are often employed in pharmacokinetic modeling. Here we report a new 8-parameter empirical mathematical model (EMM) that fits the AIF measured directly from the external femoral artery after a dose of contrast agent that was greatly reduced to minimize artifacts. The results showed that the EMM-AIF accurately models both 1^st and 2^nd passes of contrast agent circulations.

Qing Yuan¹, Daniel N Costa^1,2, Julien Sénégas³, Yin Xi¹, Andrea J Wiethoff^2,4, Robert E Lenkinski^1,2, and Ivan Pedrosa^1,2
1Radiology, UT Southwestern Medical Center, Dallas, TX, United States, ²Advanced Imaging Research Center, UT Southwestern Medical Center, Dallas, TX, United States, ³Philips Research Laboratories, Hamburg, Germany, ⁴Philips Research North America, Cambridge, MA, United States

We investigated the use of quantitative DWI and DCE measurements in MRI-visible index lesions as a surrogate for aggressiveness in prostate cancer patients. Tissue diffusion coefficient from simplified intravoxel incoherent motion model from DWI, and initial area under the curve from DCE offered the best performance in discriminating low and intermediate-to-high risk tumors. Anatomic and functional multiparametric MRI may provide a more reliable assessment of the aggressiveness of prostate cancer in patients.

Amy L. Kaczmarowski¹, Kenneth Iczkowski², William A. Hall³, Ahmad M. El-Arabi⁴, Kenneth Jacobsohn⁴, Paul Knechtges¹, Mark Hohenwalter¹, William See⁴, and Peter S. LaViolette^1
1Radiology, Medical College of Wisconsin, Milwaukee, WI, United States, ²Pathology, Medical College of Wisconsin, Milwaukee, WI, United States, ³Department of Radiation Oncology, Medical College of Wisconsin, Milwaukee, WI, United States, ⁴Urology, Medical College of Wisconsin, Milwaukee, WI, United States

Radiological-pathological correlation is being used to validate prostate cancer imaging technology. This study combines these two modalities with machine learning to generate predictive maps of histological features (i.e. new contrasts) based on segmented histology. We find that epithelium density maps highlight regions pathologically confirmed as Gleason grade ≥3. This allowed the prediction of prostate cancer presence based solely on non-invasive imaging in 23 of 26 cases.

Shiwen Shen^1,2, Xinran Zhong^1,3, Willam Hsu¹, Alex Bui¹, Holden Wu¹, Michael Kuo¹, Steven Raman¹, Daniel Margolis¹, and Kyunghyun Sung^1
1Department of Radiological Sciences, University of California, Los Angeles, Los Angeles, CA, United States, ²Department of Bioengineering, University of California, Los Angeles, Los Angeles, CA, United States,³Physics and Biology in Medicine IDP, University of California, Los Angeles, Los Angeles, CA, United States

We present a novel automatic classification method to distinguish between indolent and clinically significant prostatic carcinoma using multi-parametric MRI (mp-MRI). The main contributions are 1) utilizing state-of-art deep learning method to characterize the lesion in mp-MRI through a pre-trained convolutional neural network model, OverFeat, 2) building a hybrid two-order classification model that combines deep learning and conventional statistical features, and 3) avoiding annotation of the lesion boundaries and anatomical-location-specific training. The proposed method was evaluated using 102 lesions of prostate cancer and achieved significantly higher accuracy than the method with traditional statistical features.

Silke Hey¹, Vijayasarathy Elanchezhian², and Marius van Meel^2
1Clinical Excellence & Research, Philips HealthTech, Best, Netherlands, ²MR Clinical Applications, Philips HealthTech, Best, Netherlands

A T1w TSE Dixon acquisition is combined with view angle tilting (VAT) in order to reduce susceptibility induced artifacts from orthopedic implants close to the prostate and at the same time improve fat suppression in the area of interest. The comparison with SPIR fat suppression shows clear improvement when using Dixon together with VAT by providing more homogeneous and complete fat suppression and reduced susceptibility artifacts thus allowing clear visualization of T1 based contrast changes in the prostate and the surrounding tissue. Those results have been proven at 1.5T and 3.0T on healthy volunteers with orthopedic hip implants.

Edward William Johnston¹, Arash Latifoltojar¹, Harbir Singh Sidhu¹, Navin Ramachandran¹, Magdalena Sokolska², Alan Bainbridge², Caroline Moore³, Hashim Ahmed³, and Shonit Punwani^1
1UCL Centre for Medical Imaging, London, United Kingdom, ²Medical Physics, University College London Hospital, London, United Kingdom, ³Department of Urology, University College Hospital, London, United Kingdom

Whilst whole body MRI is gaining momentum in cancer staging for multiple tumour types, relatively few groups have focused on the primary staging of prostate cancer.  

In this study, we evaluated the role of an extensive multiparametric MRI protocol, including diffusion-weighted imaging in 23 patients against an 18F-choline PET-CT/ expert panel based reference standard.  

According to the reference standard, we found that whole body MRI provided an equivalently high diagnostic accuracy vs. PET-CT in lymph nodes, and outperformed PET-CT in the detection of bone lesions. However, higher technical error rates suggest MRI reporting experience needs to be developed first.