Prostate Gland Segmentation in Histology Images via Residual and Multi-Resolution U-Net

TL;DR

This paper introduces a modified U-Net architecture with residual and multi-resolution blocks for prostate gland segmentation in histology images, achieving superior accuracy over previous methods and aiding early prostate cancer detection.

Contribution

The paper presents a novel U-Net based model with residual and multi-resolution features that improves gland segmentation accuracy in histology images.

Findings

Residual U-Net outperforms previous state-of-the-art methods.

Achieved an average Dice Index of 0.77.

Enhanced gland detection accuracy in prostate histology images.

Abstract

Prostate cancer is one of the most prevalent cancers worldwide. One of the key factors in reducing its mortality is based on early detection. The computer-aided diagnosis systems for this task are based on the glandular structural analysis in histology images. Hence, accurate gland detection and segmentation is crucial for a successful prediction. The methodological basis of this work is a prostate gland segmentation based on U-Net convolutional neural network architectures modified with residual and multi-resolution blocks, trained using data augmentation techniques. The residual configuration outperforms in the test subset the previous state-of-the-art approaches in an image-level comparison, reaching an average Dice Index of 0.77.