Two-Stream UNET Networks for Semantic Segmentation in Medical Images

TL;DR

This paper introduces a two-stream UNET architecture that leverages intensity and gradient flow inputs to improve medical image segmentation, addressing overfitting issues in deep CNNs with limited datasets.

Contribution

The paper proposes a novel two-stream UNET model that incorporates intensity and gradient flow inputs, enhancing segmentation performance on medical images.

Findings

Two-stream UNET outperforms single-stream models on benchmarks.

Incorporating gradient flow improves segmentation accuracy.

Model is competitive with state-of-the-art methods.

Abstract

Recent advances of semantic image segmentation greatly benefit from deeper and larger Convolutional Neural Network (CNN) models. Compared to image segmentation in the wild, properties of both medical images themselves and of existing medical datasets hinder training deeper and larger models because of overfitting. To this end, we propose a novel two-stream UNET architecture for automatic end-to-end medical image segmentation, in which intensity value and gradient vector flow (GVF) are two inputs for each stream, respectively. We demonstrate that two-stream CNNs with more low-level features greatly benefit semantic segmentation for imperfect medical image datasets. Our proposed two-stream networks are trained and evaluated on the popular medical image segmentation benchmarks, and the results are competitive with the state of the art. The code will be released soon.