Produção Nacional
Can computer-aided diagnosis assist in the identification of prostate cancer on prostate MRI? a multi-center, multi-reader investigation
2018; Impact Journals LLC; Volume: 9; Issue: 73 Linguagem: Inglês
10.18632/oncotarget.26100
ISSN1949-2553
AutoresSonia Gaur, Nathan Lay, Stephanie A. Harmon, Sreya Doddakashi, Sherif Mehralivand, Burak Argun, Tristan Barrett, Sandra Bednarova, R. Girometti, Ercan Karaarslan, Ali Rıza Kural, Aytekin Oto, Andrei S. Purysko, Tatjana Antic, Cristina Magi‐Galluzzi, Yeşim Sağlıcan, Stefano Sioletic, Anne Y. Warren, Leonardo Kayat Bittencourt, Jürgen J. Fütterer, Rajan T. Gupta, Ismail Kabakus, Yan Mee Law, Daniel Margolis, Haytham Shebel, Antonio C. Westphalen, Bradford J. Wood, Peter A. Pinto, Joanna H. Shih, Peter L. Choyke, Ronald M. Summers, Barış Türkbey,
Tópico(s)Radiomics and Machine Learning in Medical Imaging
Resumo// Sonia Gaur 1 , Nathan Lay 2 , Stephanie A. Harmon 1, 3 , Sreya Doddakashi 1 , Sherif Mehralivand 1, 4, 5 , Burak Argun 6 , Tristan Barrett 7 , Sandra Bednarova 8 , Rossanno Girometti 8 , Ercan Karaarslan 9 , Ali Riza Kural 6 , Aytekin Oto 10 , Andrei S. Purysko 11 , Tatjana Antic 12 , Cristina Magi-Galluzzi 13 , Yesim Saglican 14 , Stefano Sioletic 15 , Anne Y. Warren 16 , Leonardo Bittencourt 17 , Jurgen J. Fütterer 18 , Rajan T. Gupta 19 , Ismail Kabakus 20 , Yan Mee Law 21 , Daniel J. Margolis 22 , Haytham Shebel 23 , Antonio C. Westphalen 24 , Bradford J. Wood 25 , Peter A. Pinto 4 , Joanna H. Shih 26 , Peter L. Choyke 1 , Ronald M. Summers 2 and Baris Turkbey 1 1 Molecular Imaging Program, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA 2 Imaging Biomarkers and Computer-aided Diagnosis Lab, Radiology and Imaging Sciences, Clinical Center, National Institutes of Health, Bethesda, MD, USA 3 Clinical Research Directorate/ Clinical Monitoring Research Program, Leidos Biomedical Research, Inc., Frederick National Laboratory for Cancer Research, Frederick, MD, USA 4 Urologic Oncology Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA 5 Department of Urology and Pediatric Urology, University Medical Center Mainz, Mainz, Germany 6 Department of Urology, Acibadem University, Istanbul, Turkey 7 Department of Radiology, University of Cambridge, Cambridge, UK 8 Department of Radiology, University of Udine, Udine, Italy 9 Department of Radiology, Acibadem University, Istanbul, Turkey 10 Department of Radiology, University of Chicago, Chicago, IL, USA 11 Department of Radiology, Cleveland Clinic, Cleveland, OH, USA 12 Department of Pathology, University of Chicago, Chicago, IL, USA 13 Department of Pathology, Cleveland Clinic, Cleveland, OH, USA 14 Department of Pathology, Acibadem University, Istanbul, Turkey 15 Department of Pathology, University of Udine, Udine, Italy 16 Department of Pathology, University of Cambridge, Cambridge, UK 17 Department of Radiology, Federal Fluminense University, Rio de Janeiro, Brazil 18 Department of Radiology, Radboud University, Nijmegen, The Netherlands 19 Department of Radiology, Duke University, Durham, NC, USA 20 Department of Radiology, Hacettepe University, Ankara, Turkey 21 Department of Radiology, Singapore General Hospital, Singapore 22 Weill Cornell Imaging, Cornell University, New York, NY, USA 23 Department of Radiology, Mansoura University, Mansoura, Egypt 24 UCSF Department of Radiology, University of California-San Francisco, San Francisco, CA, USA 25 Center for Interventional Oncology, Clinical Center, National Institutes of Health, Bethesda, MD, USA 26 Biometric Research Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Correspondence to: Baris Turkbey, email: ismail.turkbey@nih.gov Keywords: computer-aided diagnosis; prostate cancer; multiparametric MRI; PI-RADSv2; tumor detection Received: May 28, 2018 Accepted: August 23, 2018 Published: September 18, 2018 ABSTRACT For prostate cancer detection on prostate multiparametric MRI (mpMRI), the Prostate Imaging-Reporting and Data System version 2 (PI-RADSv2) and computer-aided diagnosis (CAD) systems aim to widely improve standardization across radiologists and centers. Our goal was to evaluate CAD assistance in prostate cancer detection compared with conventional mpMRI interpretation in a diverse dataset acquired from five institutions tested by nine readers of varying experience levels, in total representing 14 globally spread institutions. Index lesion sensitivities of mpMRI-alone were 79% (whole prostate (WP)), 84% (peripheral zone (PZ)), 71% (transition zone (TZ)), similar to CAD at 76% (WP, p=0.39), 77% (PZ, p=0.07), 79% (TZ, p=0.15). Greatest CAD benefit was in TZ for moderately-experienced readers at PI-RADSv2 <3 (84% vs mpMRI-alone 67%, p=0.055). Detection agreement was unchanged but CAD-assisted read times improved (4.6 vs 3.4 minutes, p<0.001). At PI-RADSv2 ≥ 3, CAD improved patient-level specificity (72%) compared to mpMRI-alone (45%, p<0.001). PI-RADSv2 and CAD-assisted mpMRI interpretations have similar sensitivities across multiple sites and readers while CAD has potential to improve specificity and moderately-experienced radiologists' detection of more difficult tumors in the center of the gland. The multi-institutional evidence provided is essential to future prostate MRI and CAD development.
Referência(s)