Fully Automated Prostate Segmentation on MRI: Comparison With Manual Segmentation Methods and Specimen Volumes

Abstract

Prostate volume estimates obtained with a fully automated 3D segmentation tool based on normalized gradient fields cross-correlation and graph-search refinement can yield highly accurate prostate volume estimates in a clinically relevant time of 10 seconds. This tool will assist in developing a broad range of applications including routine prostate volume estimations, image registration, biopsy guidance, and decision support systems.

Development and clinical implementation of SeedNet: A sliding‐window convolutional neural network for radioactive seed identification in MRI‐assisted radiosurgery (MARS)

“…For limiting the image context to the location of the implanted seeds, a rectangular cuboid region of interest (ROI) surrounding the prostate was cropped from the 3D image. The ROI can either be drawn manually, which was the method performed in this study, or automatically defined with one of a number of automated prostate segmentation algorithms for MRI . A sliding‐window algorithm was written to scan the entire ROI in 3D sub‐windows with a user‐defined stride.…”

MRI‐based prostate volume‐adjusted prostate‐specific antigen in the diagnosis of prostate cancer

“…This method was used in this study for analysis of the patients diagnosed with PCa. For logistic reasons, the total prostate and CG volumes of the non‐PCa patients were estimated with different automated prostate segmentation software . However, both methods were validated using prostatectomy specimens as the reference .…”

“…29 However, both methods were validated using prostatectomy specimens as the reference. 26,29 For all patients, the total prostate and the CG volumes were estimated by multiplying MRI voxel size (in cm 3 ) by the number of voxels within a segmented volume. Subsequently, the PZ volume was estimated by subtracting the CG volume from the total prostate volume.…”

Diagnostic value of MRI-based PSA density in predicting transperineal sector-guided prostate biopsy outcomes

“…This calculation does require user intervention in measuring the prostate dimensions and is thus subject to inter-observer variation [4,5]. A potential solution lies in MRI segmentation for prostate volume calculation; fully-automated methods are able to yield highly accurate volumes, and reduce this user variation [40]. Despite this limitation, error was reduced by applying the same volume calculation across all patients; all prostate dimensions were calculated by senior uroradiologists.…”