Pulse Sequence Dependence of a Simple and Interpretable Deep Learning Method for Detection of Clinically Significant Prostate Cancer Using Multiparametric MRI

•A simple convolutional network incorporating 3 pulse-sequences outperforms any one sequence for detecting significant prostate cancer. •A composite including dynamic contrast enhancement correlates with grade. •This model may outperform conventional assessment methods, such as PI-RADS. Rationale and Objectives Multiparametric magnetic resonance imaging (mpMRI) is increasingly used for risk stratification and localization of prostate cancer (PCa). Thanks to the great success of deep learning models in computer vision, the potential application for early detection of PCa using mpMRI is imminent. Materials and Methods Deep learning analysis of the PROSTATEx dataset. Results In this study, we show a simple convolutional neural network (CNN) with mpMRI can achieve high performance for detection of clinically significant PCa (csPCa), depending on the pulse sequences used. The mpMRI model with T2-ADC-DWI achieved 0.90 AUC score in the held-out test set, not significantly better than the model using Ktrans instead of DWI (AUC 0.89). Interestingly, the model incorporating T2-ADC- Ktrans better estimates grade. We also describe a saliency “heat” map. Our results show that csPCa detection models with mpMRI may be leveraged to guide clinical management strategies. Conclusion Convolutional neural networks incorporating multiple pulse sequences show high performance for detection of clinically-significant prostate cancer, and the model including dynamic contrast-enhanced information correlates best with grade. Multiparametric magnetic resonance imaging (mpMRI) is increasingly used for risk stratification and localization of prostate cancer (PCa). Thanks to the great success of deep learning models in computer vision, the potential application for early detection of PCa using mpMRI is imminent. Deep learning analysis of the PROSTATEx dataset. In this study, we show a simple convolutional neural network (CNN) with mpMRI can achieve high performance for detection of clinically significant PCa (csPCa), depending on the pulse sequences used. The mpMRI model with T2-ADC-DWI achieved 0.90 AUC score in the held-out test set, not significantly better than the model using Ktrans instead of DWI (AUC 0.89). Interestingly, the model incorporating T2-ADC- Ktrans better estimates grade. We also describe a saliency “heat” map. Our results show that csPCa detection models with mpMRI may be leveraged to guide clinical management strategies. Convolutional neural networks incorporating multiple pulse sequences show high performance for detection of clinically-significant prostate cancer, and the model including dynamic contrast-enhanced information correlates best with grade.