Abstract
Plane-wave ultrasound imaging offers advantages in terms of frame rate and penetration depth, but often suffers from reduced image quality compared to focused ultrasound. Super-resolution reconstruction techniques can enhance image quality, but traditional approaches often fail to preserve fine details and anatomical structures. In this work, we propose a novel multi-angle parallel U-Net architecture with maxout units for ultrasound image super-resolution.
Our approach introduces a parallel processing framework that handles multiple imaging angles simultaneously. The framework consists of three key components: (1) a multi-angle parallel U-Net that processes different viewing angles, (2) maxout units that enhance feature representation, and (3) a novel loss function that preserves both structural and textural details.
We evaluate our method on plane-wave ultrasound datasets with ground truth high-resolution images. Experimental results demonstrate that our approach significantly improves image quality compared to existing super-resolution methods. The method shows excellent performance in preserving fine anatomical details and reducing artifacts commonly found in ultrasound imaging.
The proposed framework represents a significant advancement in ultrasound image processing, providing higher resolution images that could improve diagnostic accuracy and clinical decision-making.
BibTeX
@article{zhou2019super,
title={Super-Resolution Reconstruction of Plane-Wave Ultrasound Image Based on a Multi-Angle Parallel U-Net with Maxout Unit and Novel Loss Function},
author={Zhou, Zixia and Wang, Yuanyuan and Yu, Jinhua and Guo, Wei and Li, Zeju},
journal={Journal of Medical Imaging and Health Informatics},
year={2019},
publisher={American Scientific Publishers},
doi={10.1166/jmihi.2019.2548}
}