Dataset underlying the study: Waveform-Specific Performance of Deep Learning-Based Super-Resolution for Ultrasound Contrast Imaging

This dataset contains the data used for the study ‘Waveform-Specific Performance of Deep Learning-Based Super-Resolution for Ultrasound Contrast Imaging’ (link will be added after publication). <strong>Contents:</strong> It consists of radiofrequency (RF) signals acquired during simulations and experiments, weights and biases of networks trained, and images reconstructed using delay-and-sum (DAS) beamforming. In addition to that, it also contains environments and packages used to process the data. <strong>Objective:</strong> To investigate the effect of transmit waveforms on the performance of deep learning-based approaches for localizing microbubbles in radiofrequency signals. <strong>Type of research:</strong> Fundamental, Physics, Biomedical. <strong>Method of data collection:</strong> In-silico and in-vitro. <strong>Type of data:</strong> RF signals (<code style="color:rgb(199, 37, 78);background-color:rgb(249, 242, 244);">.mat</code> and <code style="color:rgb(199, 37, 78);background-color:rgb(249, 242, 244);">.txt</code>), super-resolved RF signals (<code style="color:rgb(199, 37, 78);background-color:rgb(249, 242, 244);">.txt</code> and <code style="color:rgb(199, 37, 78);background-color:rgb(249, 242, 244);">.npy</code>), weights and biases (<code style="color:rgb(199, 37, 78);background-color:rgb(249, 242, 244);">.txt</code>), images generated with the (super-resolved) RF signals (<code style="color:rgb(199, 37, 78);background-color:rgb(249, 242, 244);">.mat</code>), videos generated with the (super-resolved) RF signals (<code style="color:rgb(199, 37, 78);background-color:rgb(249, 242, 244);">.mp4</code>), python environments (<code style="color:rgb(199, 37, 78);background-color:rgb(249, 242, 244);">.yaml</code>), microbubble size distributions (<code style="color:rgb(199, 37, 78);background-color:rgb(249, 242, 244);">.fig</code> and <code style="color:rgb(199, 37, 78);background-color:rgb(249, 242, 244);">.png</code>). The code used for this work is available at https://github.com/MIAGroupUT/super-resolution-waveforms.本数据集用于支撑题为《超声造影成像领域基于深度学习的超分辨率技术的波形特异性性能》的研究（论文正式发表后将补充链接）。 **数据集内容：** 本数据集涵盖仿真与实验场景中采集的射频（Radiofrequency, RF）信号、训练所得的网络权重与偏置，以及通过延时叠加（Delay-and-Sum, DAS）波束形成算法重建的图像。此外，本数据集还包含数据处理所需的运行环境与依赖包。 **研究目标：** 探究发射波形对基于深度学习的射频信号微泡定位方法性能的影响。 **研究类型：** 基础物理学、生物医学领域研究。 **数据采集方式：** 计算机仿真（In-silico）与体外实验（In-vitro）。 **数据类型：** 具体包含以下格式的多类数据： 1. 射频信号（格式为`.mat`与`.txt`）； 2. 超分辨率射频信号（格式为`.txt`与`.npy`）； 3. 网络权重与偏置文件（格式为`.txt`）； 4. 基于（超分辨率）射频信号生成的重建图像（格式为`.mat`）； 5. 基于（超分辨率）射频信号生成的可视化视频（格式为`.mp4`）； 6. Python运行环境配置文件（格式为`.yaml`）； 7. 微泡尺寸分布相关数据（格式为`.fig`与`.png`）。 本研究所用代码已开源，开源地址为：https://github.com/MIAGroupUT/super-resolution-waveforms。