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

Abstract: We present a novel multi-stage 3D computer-aided detection and diagnosis (CAD) model for automated localization of clinically significant prostate cancer (csPCa) in bi-parametric MR imaging (bpMRI). State-of-the-art attention mechanisms drive its detection network, which aims to accurately discriminate csPCa lesions from indolent cancer and the wide range of benign pathology that can afflict the prostate gland. In parallel, a decoupled residual classifier is used to achieve consistent false positive reduction, without sacrificing high detection sensitivity or computational efficiency. Furthermore, a probabilistic anatomical prior, which captures the spatial prevalence of csPCa and its zonal distinction, is computed and encoded into the CNN architecture to guide model generalization with domain-specific clinical knowledge.
For 486 institutional testing scans, the 3D CAD system achieves $83.69\pm5.22\%$ and $93.19\pm2.96\%$ detection sensitivity at 0.50 and 1.46 false positive(s) per patient, respectively, along with $0.882$ AUROC in patient-based diagnosis $-$significantly outperforming four state-of-the-art baseline architectures (USEResNet, UNet++, nnU-Net, Attention U-Net) from recent literature. For 296 external testing scans, the ensembled CAD system shares moderate agreement with a consensus of expert radiologists ($76.69\%$; $kappa=0.511$) and independent pathologists ($81.08\%$; $kappa=0.559$); demonstrating a strong ability to localize histologically-confirmed malignancies and generalize beyond the radiologically-estimated annotations of the 1950 training-validation cases used in this study.