Aritrick Chatterjee¹, Carla Harmath¹, Roger Engelmann¹, Ajit Devaraj², Ambereen Yousuf¹, Scott Eggener³, Glenn Gerber³, Gregory Karczmar¹, and Aytekin Oto^1
1Department of Radiology, University of Chicago, Chicago, IL, United States, ²Philips Research North America, Chicago, IL, United States, ³Department of Urology, University of Chicago, Chicago, IL, United States

This prospective clinical trial evaluates whether HM-MRI identifies PCa more reliably than random biopsy and/or targets detected based on PI-RADSv2. Patients underwent 3T mpMRI along with HM-MRI. Patients received 12-core TRUS-guided sextant random biopsy. Additional biopsy targets selected by radiologist (≥PI-RADS 3) and suspected PCa based on HM-MRI tissue composition estimates were biopsied, using a Uronav MR-US fusion biopsy device. The diagnostic accuracy of HM-MRI for detecting clinically significant cancers was higher than that of mpMRI on per-tumor (0.74 vs 0.61) and sextant analysis (0.84 vs 0.75). HM-MRI had higher accuracy, sensitivity, specificity and PPV than mpMRI, with similar NPV.

Ruiming Cao¹, Xinran Zhong², Sohrab Afshari Mirak³, Ely Felker³, Voraparee Suvannarerg^3,4, Teeravut Tubtawee^3,5, Fabien Scalzo⁶, Steve Raman³, and Kyunghyun Sung^3
1Bioengineering, UC Berkeley, Berkeley, CA, United States, ²UT Southwestern, Dallas, TX, United States, ³Radiology, UCLA, Los Angeles, CA, United States, ⁴Radiology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand, ⁵Radiology, Faculty of Medicine, Prince of Songkla University, Songkhla, Thailand, ⁶Computer Science, UCLA, Los Angeles, CA, United States

We evaluated our recently developed deep learning system, FocalNet, for prostate cancer detection in multi-parametric MRI (mpMRI). This study performed a head-to-head comparison between FocalNet and four genitourinary radiologists in an independent evaluation cohort consisting of 126 mpMRI scans untouched during the development. FocalNet demonstrated similar detection performance to radiologists under the high specificity condition or the high sensitivity condition, while radiologists outperformed FocalNet in moderate specificity and sensitivity.

Jonathan J. Sackett¹, Joanna Shih¹, Sarah Reese¹, Jeffrey R. Brender^1,2, Stephanie Harmon¹, Tristan Barrett¹, Mehmet Coskun¹, Manuel Madariaga ¹, Jamie Marko ¹, Yan Mee Law ¹, Evrim Turkbey ¹, Sherif Mehralivand ¹, Thomas Sanford ¹, Nathan Lay ¹, Peter A. Pinto ¹, Bradford J. Wood², Peter L. Choyke², Murali C. Krishna², and Baris I Turkbey^2
1Molecular Imaging Program, National Cancer Institute, National Institutes of Health, Bethesda, MD, MD, United States, ²National Cancer Institute, National Institutes of Health, Bethesda, MD, MD, United States

High quality MRIs are needed for prostate cancer screening and accurate targeting in MRI guided biopsies. Based on blind image quality assessment by image readers, compliance with the PIRADSv2 Minimum Technical Standards was determined to be neither necessary nor sufficient in ensuring quality in prostate T2 and DWI images. 

Sean D McGarry¹, John D Bukowy², Kenneth Iczkowsk³, Allison K Lowman², Michael Brehler², Samuel Bobholz¹, Alex Barrington², Kenneth Jacobsohn⁴, Jackson Unteriner², Petar Duvnjak², Michael Griffin², Mark Hohenwalter², Tucker Keuter⁵, Wei Huang⁶, Tatjana Antic⁷, Gladell Paner⁷, Watchareepohn Palanghmonthip^3,8, Anjishnu Banerjee⁵, and Peter S LaViolette^2
1Biophysics, Medical College of Wisconsin, Milwaukee, WI, United States, ²Radiology, Medical College of Wisconsin, Milwaukee, WI, United States, ³Pathology, Medical College of Wisconsin, Milwaukee, WI, United States, ⁴Urological Surgery, Medical College of Wisconsin, Milwaukee, WI, United States, ⁵Biostatistics, Medical College of Wisconsin, Milwaukee, WI, United States, ⁶Pathology, University of Wisconsin Madison, Madison, WI, United States, ⁷Pathology, University of Chicago, Chicago, IL, United States, ⁸Pathology, Chiang Mai University, Chiang Mai, Thailand

This study demonstrates that radio-pathomic maps of epithelium density derived from annotations performed by different pathologists distinguish high-grade prostate cancer from G3 and benign atrophy. In a test set of 5 patients epithelium density maps consistently demonstrate an AUC greater than 0.9 independent of which pathologist’s annotations trained the model or which pathologist’s annotations the model is applied to. The results in a larger test set largely mirror the results in the small test set. We also showed that radio-pathomic maps of epithelium density out-performed ADC maps independent of which observer was used to train the model.

Zhaohuan Zhang^1,2 and Holden Wu^1,2
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

Recent work has demonstrated the ability of Diffusion-Relaxation Correlation Spectrum Imaging (DR-CSI) to quantify prostate microscopic tissue compartments (epithelium, stroma and lumen) in ex vivo prostate specimens at 3T using whole-mount digital histopathology as the reference. This study further developed DR-CSI for in vivo characterization of prostate microstructure using high resolution ex vivo DR-CSI as the reference. Consistent trends in DR-CSI signal component fraction variations in different prostate regions were found using in vivo DR-CSI across subjects, and agreed with trends on ex vivo DR-CSI. 

Christopher C Conlin¹, Christine H Feng², Ana E Rodriguez-Soto¹, Roshan A Karunamuni², Joshua M Kuperman¹, Dominic Holland¹, Rebecca Rakow-Penner¹, Tyler M Seibert², Anders M Dale^1,3, and Michael E Hahn^1
1Department of Radiology, UC San Diego School of Medicine, La Jolla, CA, United States, ²Department of Radiation Medicine and Applied Science, UC San Diego School of Medicine, La Jolla, CA, United States, ³Department of Neurosciences, UC San Diego School of Medicine, La Jolla, CA, United States

Restriction spectrum imaging (RSI) is an advanced multi-shell diffusion technique that models the diffusion-weighted signal as a linear combination of exponential decays. While RSI shows promise for assessing prostate cancer, an optimal RSI model that effectively characterizes the diffusion properties of both normal and cancerous prostate tissue is essential to ensuring an accurate evaluation of prostate cancer lesions. In this study, we determined optimal ADC values for several RSI models of the prostate and assessed the number of tissue compartments required to best describe diffusion in both normal and cancerous prostate tissue.

Shirin Sabouri¹, Silvia Chang^2,3,4, Emily Pang², Rehab Mohammedeid², Edward Jones⁵, Larry Goldenberg^3,4, Peter Black^3,4, and Piotr Kozlowski^1,2,3,4
1UBC MRI Research Centre, Vancouver, BC, Canada, ²Department of Radiology, The University of British Columbia, Vancouver, BC, Canada, ³Department of Urologic Sciences, The University of British Columbia, Vancouver, BC, Canada, ⁴Vancouver Prostate Centre, Vancouver, BC, Canada, ⁵Department of Pathology and Laboratory Medicine, The University of British Columbia, Vancouver, BC, Canada

Luminal water imaging (LWI) is an MRI technique that detects regions of decreased lumen volume in prostate. Recent studies on LWI have shown promising results regarding the accuracy of this technique in diagnosis of prostate cancer. However, to the best of our knowledge no study has yet compared the performance of LWI against the current clinical assessment. In this study, we perform a comparison between the diagnostic accuracy of LWI with the PI-RADSv2.1 assessment. Our results show that LWI performs similar to the PI-RADSv2.1 in the entire prostate and peripheral zone, and outperforms significantly in transition zone.

Marco Palombo¹, Saurabh Singh², Hayley Whitaker², Shonit Puwani², Daniel C. Alexander¹, and Eleftheria Panagiotaki^1
1Centre for Medical Image Computing, Department of Computer Science, University College London, London, United Kingdom, ²Centre for Medical Imaging, University College London, London, United Kingdom

This work presents a comprehensive VERDICT prostate model (relaxed-VERDICT) that includes compartment-specific relaxation effects providing prostate microstructural estimates unbiased by the relaxation properties of the tissue. We compare relaxed-VERDICT with the original VERDICT model and use it to provide estimates of T2 and T1 in benign and tumor prostate tissue. Our results suggest that original VERDICT’s f_ic contrast is mostly driven by diffusion, supporting its use as imaging marker of apparent cellular volume fraction. Relaxed-VERDICT estimates of T1/T2 values are in very good agreement with literature. Finally, we propose a machine learning based processing pipeline that provides ultra-fast quantitative maps.

Ethan Leng¹, Benjamin Spilseth², Anil Chauhan², Joseph Gill², Ana Rosa², Arcan Erturk¹, Naoharu Kobayashi¹, Xiaoxuan He¹, Christopher Warlick³, and Gregory Metzger^1
1Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States, ²Radiology, University of Minnesota, Minneapolis, MN, United States, ³Urology, University of Minnesota, Minneapolis, MN, United States

Recent works have demonstrated the feasibility of prostate multiparametric MRI (mpMRI) at 7T with improved resolution compared to mpMRI at 3T. However, the clinical relevance of finer anatomic details versus the drawbacks of increased imaging artifacts at 7T has yet to be investigated. In this work, we conducted a retrospective, multi-reader clinical evaluation of 19 paired mpMRI studies at 3T and 7T. The primary outcome of interest was accuracy of prostate cancer detection, with image quality and artifacts as secondary outcomes.