Segmentation of the Prostatic Gland and the Intraprostatic Lesions on Multiparametic MRI Using Mask R-CNN

Abstract Background and Purpose Accurate delineation of the prostate gland and intraprostatic lesions (ILs) is essential for prostate cancer (PCa) dose-escalated radiation therapy. The aim of this study was to develop a sophisticated deep neural network approach to magnetic resonance (MR) image analysis that will help IL detection and delineation for clinicians. Materials and Methods We trained and evaluated mask region-based convolutional neural networks (Mask R-CNNs) to perform the prostate gland and IL segmentation. There were two cohorts in this study: 78 public patients (cohort 1) and 42 private patients from our institution (cohort 2). Prostate gland segmentation was performed using T2 weighted images (T2WIs), while IL segmentation was performed using T2WIs and co-registered apparent diffusion coefficient (ADC) maps with prostate patches cropped out. The IL segmentation model was extended to select five highly suspicious volumetric lesions within the entire prostate. Results The Mask R-CNN model was able to segment the prostate with dice similarity coefficient (DSC) of 0.88 ± 0.04, 0.86 ± 0.04, 0.82 ± 0.05, sensitivity (Sens.) of 0.93, 0.95, 0.95 and specificity (Spec.) of 0.98, 0.85, 0.90; while ILs were segmented with DSC of 0.62 ± 0.17, 0.59 ± 0.14, 0.38 ± 0.19, Sens. of 0.55 ± 0.30, 0.63 ± 0.28, 0.22 ± 0.24 and Spec. of 0.974 ± 0.010, 0.964 ± 0.015, 0.972 ± 0.015 in public validation/public testing/private testing patients when trained with patients from cohort 1 only. When trained with patients from both cohorts, the values were as follows: DSC of 0.64 ± 0.11, 0.56 ± 0.15, 0.46 ± 0.15, Sens. of 0.57 ± 0.23, 0.50 ± 0.28, 0.33 ± 0.17 and Spec. of 0.980 ± 0.009, 0.969 ± 0.016, 0.977 ± 0.013. Conclusion Our research framework is able to perform as an end-to-end system that automatically segmented prostate gland, identified and delineated highly suspicious ILs within the entire prostate. Therefore, this system demonstrated the potential for assisting the clinicians in tumor delineation.