Prostate Cancer Semantic Segmentation by Gleason Score Group in bi-parametric MRI with Self Attention Model on the Peripheral Zone
Proceedings of the Third Conference on Medical Imaging with Deep Learning, PMLR 121:193-204, 2020.
In this work, we propose a novel end-to-end multi-class attention network to jointly perform peripheral zone (PZ) segmentation and PZ lesions detection with Gleason score (GS) group grading. After encoding the information on a latent space, the network is separated in two branches: 1) the first branch performs PZ segmentation 2) the second branch uses this zonal prior as an attention gate for the detection and grading of PZ lesions. The model was trained and validated with a 5-fold cross-validation on an heterogeneous series of 98 MRI exams acquired on two different scanners prior prostatectomy. In the free-response receiver operating characteristics (FROC) analysis for clinically significant lesions (defined as GS $> 6$) detection, our model achieves $75.8% \pm 3.4$% sensitivity at 2.5 false positive per patient. Regarding the automatic GS group grading, Cohen’s quadratic weighted kappa coefficient is $0.35 \pm 0.05$, which is considered as a fair agreement and an improvement with regards to the baseline U-Net model. Our method achieves good performance without requiring any prior manual region delineation in clinical practice. We show that the addition of the attention mechanism improves the CAD performance in comparison to the baseline model.