End-to-end Prostate Cancer Detection in bpMRI via 3D CNNs: Effect of Attention Mechanisms, Clinical Priori and Decoupled False Positive Reduction

We present a multi-stage 3D computer-aided detection and diagnosis (CAD) model2 for automated localization of clinically significant prostate cancer (csPCa) in bi-parametric MR imaging (bpMRI). Deep attention mechanisms drive its detection network, targeting salient structures and highly discriminative feature dimensions across multiple resolutions. Its goal is to accurately identify csPCa lesions from indolent cancer and the wide range of benign pathology that can afflict the prostate gland. Simultaneously, a decoupled residual classifier is used to achieve consistent false positive reduction, without sacrificing high sensitivity or computational efficiency. In order to guide model generalization with domain-specific clinical knowledge, a probabilistic anatomical prior is used to encode the spatial prevalence and zonal distinction of csPCa. Using a large dataset of 1950 prostate bpMRI paired with radiologically-estimated annotations, we hypothesize that such CNN-based models can be trained to detect biopsy-confirmed malignancies in an independent cohort. For 486 institutional testing scans, the 3D CAD system achieves 83.69±5.22% and 93.19±2.96% detection sensitivity at 0.50 and 1.46 false positive(s) per patient, respectively, with 0.882±0.030 AUROC in patient-based diagnosis -significantly outperforming four state-of-the-art baseline architectures (U-SEResNet, UNet++, nnU-Net, Attention U-Net) from recent literature. For 296 external biopsy-confirmed testing scans, the ensembled CAD system shares moderate agreement with a consensus of expert radiologists (76.69%; kappa = 0.51±0.04) and independent pathologists (81.08%; kappa = 0.56±0.06); demonstrating strong generalization to histologically-confirmed csPCa diagnosis.

[1]  Giancarlo Mauri,et al.  USE-Net: incorporating Squeeze-and-Excitation blocks into U-Net for prostate zonal segmentation of multi-institutional MRI datasets , 2019, Neurocomputing.

[2]  C. Tempany,et al.  The Current State of MR Imaging-targeted Biopsy Techniques for Detection of Prostate Cancer. , 2017, Radiology.

[3]  Laura M. Chihara,et al.  Mathematical Statistics with Resampling and R , 2011 .

[4]  Josien P. W. Pluim,et al.  Not‐so‐supervised: A survey of semi‐supervised, multi‐instance, and transfer learning in medical image analysis , 2018, Medical Image Anal..

[5]  Ivan Pedrosa,et al.  Geometric distortion in diffusion-weighted MR imaging of the prostate-contributing factors and strategies for improvement. , 2014, Academic radiology.

[6]  H. Huisman,et al.  Encoding Clinical Priori in 3D Convolutional Neural Networks for Prostate Cancer Detection in bpMRI , 2020, ArXiv.

[7]  D. Margolis,et al.  PI-RADS Prostate Imaging - Reporting and Data System: 2015, Version 2. , 2016, European urology.

[8]  Masoom A. Haider,et al.  Prostate Cancer Detection via a Quantitative Radiomics-Driven Conditional Random Field Framework , 2015, IEEE Access.

[9]  Manuel Wiesenfarth,et al.  Classification of Cancer at Prostate MRI: Deep Learning versus Clinical PI-RADS Assessment. , 2019, Radiology.

[10]  Nico Karssemeijer,et al.  Computer-Aided Detection of Prostate Cancer in MRI , 2014, IEEE Transactions on Medical Imaging.

[11]  Mert R. Sabuncu,et al.  Anatomical Priors in Convolutional Networks for Unsupervised Biomedical Segmentation , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[12]  Thomas Brox,et al.  3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation , 2016, MICCAI.

[13]  Fabien Scalzo,et al.  Prostate Cancer Inference via Weakly-Supervised Learning using a Large Collection of Negative MRI , 2019, 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW).

[14]  Enhua Wu,et al.  Squeeze-and-Excitation Networks , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[15]  Jeroen van der Laak,et al.  Tailoring automated data augmentation to H&E-stained histopathology , 2021, MIDL.

[16]  Ben Glocker,et al.  Attention Gated Networks: Learning to Leverage Salient Regions in Medical Images , 2018, Medical Image Anal..

[17]  Tayyar A. Ozkan,et al.  Interobserver variability in Gleason histological grading of prostate cancer , 2016, Scandinavian journal of urology.

[18]  D. Johnston,et al.  Detailed mapping of prostate carcinoma foci , 2000, Cancer.

[19]  B. Delahunt,et al.  The 2014 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma: Definition of Grading Patterns and Proposal for a New Grading System , 2015, The American journal of surgical pathology.

[20]  Ruiming Cao,et al.  Joint Prostate Cancer Detection and Gleason Score Prediction in mp-MRI via FocalNet , 2019, IEEE Transactions on Medical Imaging.

[21]  David S. Melnick,et al.  International evaluation of an AI system for breast cancer screening , 2020, Nature.

[22]  Dacheng Tao,et al.  Two-Stage Cascaded U-Net: 1st Place Solution to BraTS Challenge 2019 Segmentation Task , 2019, BrainLes@MICCAI.

[23]  Baris Turkbey,et al.  Multiparametric MRI in prostate cancer management , 2014, Nature Reviews Clinical Oncology.

[24]  Iasonas Kokkinos,et al.  Harnessing Uncertainty in Domain Adaptation for MRI Prostate Lesion Segmentation , 2020, MICCAI.

[25]  Adam T Froemming,et al.  Interobserver Reproducibility of the PI-RADS Version 2 Lexicon: A Multicenter Study of Six Experienced Prostate Radiologists. , 2016, Radiology.

[26]  Masoom A. Haider,et al.  Prostate Cancer Detection using Deep Convolutional Neural Networks , 2019, Scientific Reports.

[27]  Carole Lartizien,et al.  Prostate Cancer Semantic Segmentation by Gleason Score Group in bi-parametric MRI with Self Attention Model on the Peripheral Zone , 2020, MIDL.

[28]  Abhishek Das,et al.  Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization , 2016, 2017 IEEE International Conference on Computer Vision (ICCV).

[29]  Yongdong Zhang,et al.  A Two-Stream Mutual Attention Network for Semi-Supervised Biomedical Segmentation with Noisy Labels , 2018, AAAI.

[30]  Kingshuk Roy Choudhury,et al.  Detection of prostate cancer with multiparametric MRI (mpMRI): effect of dedicated reader education on accuracy and confidence of index and anterior cancer diagnosis , 2014, Abdominal Imaging.

[31]  Jun Fu,et al.  Dual Attention Network for Scene Segmentation , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[32]  Indrani Bhattacharya,et al.  CorrSigNet: Learning CORRelated Prostate Cancer SIGnatures from Radiology and Pathology Images for Improved Computer Aided Diagnosis , 2020, MICCAI.

[33]  Jian Sun,et al.  Identity Mappings in Deep Residual Networks , 2016, ECCV.

[34]  Yoshua Bengio,et al.  Inductive Biases for Deep Learning of Higher-Level Cognition , 2020, ArXiv.

[35]  Christina Hulsbergen-van de Kaa,et al.  Head-to-head Comparison of Transrectal Ultrasound-guided Prostate Biopsy Versus Multiparametric Prostate Resonance Imaging with Subsequent Magnetic Resonance-guided Biopsy in Biopsy-naïve Men with Elevated Prostate-specific Antigen: A Large Prospective Multicenter Clinical Study. , 2019, European urology.

[36]  P. Choyke,et al.  Deep‐Learning‐Based Artificial Intelligence for PI‐RADS Classification to Assist Multiparametric Prostate MRI Interpretation: A Development Study , 2020, Journal of magnetic resonance imaging : JMRI.

[37]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[38]  Bin Lou,et al.  False Positive Reduction Using Multiscale Contextual Features for Prostate Cancer Detection in Multi-Parametric MRI Scans , 2020, 2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI).

[39]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[40]  Fatma Taher,et al.  A 2.5D Deep Learning-Based Approach for Prostate Cancer Detection on T2-Weighted Magnetic Resonance Imaging , 2018, ECCV Workshops.

[41]  Dorin Comaniciu,et al.  Deep Attentive Panoptic Model for Prostate Cancer Detection Using Biparametric MRI Scans , 2020, MICCAI.

[42]  Laurent Lemaitre,et al.  Use of prostate systematic and targeted biopsy on the basis of multiparametric MRI in biopsy-naive patients (MRI-FIRST): a prospective, multicentre, paired diagnostic study. , 2019, The Lancet. Oncology.

[43]  Gavin C. Cawley,et al.  On Over-fitting in Model Selection and Subsequent Selection Bias in Performance Evaluation , 2010, J. Mach. Learn. Res..

[44]  Jian Sun,et al.  Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[45]  Mitko Veta,et al.  Deep Learning Regression for Prostate Cancer Detection and Grading in Bi-Parametric MRI , 2020, IEEE Transactions on Biomedical Engineering.

[46]  A. Padhani,et al.  Multiparametric Magnetic Resonance Imaging for the Detection of Clinically Significant Prostate Cancer: What Urologists Need to Know. Part 2: Interpretation. , 2019, European urology.

[47]  Xiaogang Wang,et al.  Residual Attention Network for Image Classification , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[48]  Tobias Penzkofer,et al.  Semi-automatic classification of prostate cancer on multi-parametric MR imaging using a multi-channel 3D convolutional neural network , 2019, European Radiology.

[49]  Jinjun Xiong,et al.  Revisiting RCNN: On Awakening the Classification Power of Faster RCNN , 2018, ECCV.

[50]  Peter A. Calabresi,et al.  Evaluating White Matter Lesion Segmentations with Refined Sørensen-Dice Analysis , 2020, Scientific Reports.

[51]  Julien Cohen-Adad,et al.  SoftSeg: Advantages of soft versus binary training for image segmentation , 2020, Medical Image Anal..

[52]  P. Walsh Influence of prostate volume in the detection of prostate cancer. , 2003, The Journal of urology.

[53]  Hao Tang,et al.  NoduleNet: Decoupled False Positive Reductionfor Pulmonary Nodule Detection and Segmentation , 2019, MICCAI.

[54]  J. Hanley,et al.  The meaning and use of the area under a receiver operating characteristic (ROC) curve. , 1982, Radiology.

[55]  B. Kramer,et al.  Cancer overdiagnosis: a biological challenge and clinical dilemma , 2019, Nature Reviews Cancer.

[56]  Nima Tajbakhsh,et al.  UNet++: Redesigning Skip Connections to Exploit Multiscale Features in Image Segmentation , 2020, IEEE Transactions on Medical Imaging.

[57]  A. Padhani,et al.  A systematic review and meta-analysis of the diagnostic accuracy of biparametric prostate MRI for prostate cancer in men at risk , 2020, Prostate Cancer and Prostatic Diseases.

[58]  M. Roethke,et al.  Multiparametric Magnetic Resonance Imaging (MRI) and MRI-Transrectal Ultrasound Fusion Biopsy for Index Tumor Detection: Correlation with Radical Prostatectomy Specimen. , 2016, European urology.

[59]  Baris Turkbey,et al.  Variability of the Positive Predictive Value of PI-RADS for Prostate MRI across 26 Centers: Experience of the Society of Abdominal Radiology Prostate Cancer Disease-focused Panel. , 2020, Radiology.

[60]  Sergey Ioffe,et al.  Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning , 2016, AAAI.

[61]  D. Margolis,et al.  MRI‐Targeted or Standard Biopsy for Prostate‐Cancer Diagnosis , 2018, The New England journal of medicine.

[62]  P. Whiting,et al.  Comparison of Multiparametric Magnetic Resonance Imaging and Targeted Biopsy With Systematic Biopsy Alone for the Diagnosis of Prostate Cancer , 2019, JAMA network open.

[63]  Imam Samil Yetik,et al.  Prostate Cancer Localization Using Multiparametric MRI based on Semisupervised Techniques With Automated Seed Initialization , 2012, IEEE Transactions on Information Technology in Biomedicine.

[64]  Ahmedin Jemal,et al.  Cancer treatment and survivorship statistics, 2019 , 2019, CA: a cancer journal for clinicians.

[65]  Dean C. Barratt,et al.  Automatic Multi-Organ Segmentation on Abdominal CT With Dense V-Networks , 2018, IEEE Transactions on Medical Imaging.

[66]  Mingyan Liu,et al.  Characterizing Adversarial Examples Based on Spatial Consistency Information for Semantic Segmentation , 2018, ECCV.

[67]  S. Klein,et al.  A Multi-Center, Multi-Vendor Study to Evaluate the Generalizability of a Radiomics Model for Classifying Prostate cancer: High Grade vs. Low Grade , 2021, Diagnostics.

[68]  Ross B. Girshick,et al.  Focal Loss for Dense Object Detection , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[69]  Nathan Lay,et al.  Prostate cancer detection from multi-institution multiparametric MRIs using deep convolutional neural networks , 2018, Journal of medical imaging.

[70]  A. Padhani,et al.  Multiparametric Magnetic Resonance Imaging for the Detection of Clinically Significant Prostate Cancer: What Urologists Need to Know. Part 1: Acquisition. , 2020, European urology.

[71]  Imon Banerjee,et al.  An Automated Two-step Pipeline for Aggressive Prostate Lesion Detection from Multi-parametric MR Sequence. , 2020, AMIA Joint Summits on Translational Science proceedings. AMIA Joint Summits on Translational Science.

[72]  Jens Petersen,et al.  nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation , 2020, Nature Methods.

[73]  Sebastian Thrun,et al.  Dermatologist-level classification of skin cancer with deep neural networks , 2017, Nature.

[74]  Leslie N. Smith,et al.  Cyclical Learning Rates for Training Neural Networks , 2015, 2017 IEEE Winter Conference on Applications of Computer Vision (WACV).

[75]  Baris Turkbey,et al.  Intra‐ and interreader reproducibility of PI‐RADSv2: A multireader study , 2019, Journal of magnetic resonance imaging : JMRI.

[76]  Christian Wachinger,et al.  DeepNAT: Deep convolutional neural network for segmenting neuroanatomy , 2017, NeuroImage.

[77]  Richard E. Fan,et al.  Automated detection of aggressive and indolent prostate cancer on magnetic resonance imaging , 2021, Medical physics.

[78]  D. Berney,et al.  Standardization of Gleason grading among 337 European pathologists , 2013, Histopathology.

[79]  A. Ragozzino,et al.  Assessment of DCE Utility for PCa Diagnosis Using PI-RADS v2.1: Effects on Diagnostic Accuracy and Reproducibility , 2020, Diagnostics.

[80]  Henkjan Huisman,et al.  Effect of Adding Probabilistic Zonal Prior in Deep Learning-based Prostate Cancer Detection. , 2019, 1907.12382.

[81]  D. Margolis,et al.  Prostate Imaging Reporting and Data System Version 2.1: 2019 Update of Prostate Imaging Reporting and Data System Version 2. , 2019, European urology.

[82]  Klaus H. Maier-Hein,et al.  Adversarial Networks for the Detection of Aggressive Prostate Cancer , 2017, ArXiv.

[83]  Xu Sun,et al.  Adaptive Gradient Methods with Dynamic Bound of Learning Rate , 2019, ICLR.