MedSeg Covid Dataset 1

This is a dataset of 100 axial CT images from &gt;40 patients with COVID-19 that were converted from openly accessible JPG images found at https://www.sirm.org/en/ . The conversion process is described in detail in the following blogpost: Covid-19 radiology — data collection and preparation for Artificial Intelligence<br>In short, the images were segmented by a radiologist using 3 labels: <em>ground-glass</em> (mask value =1), <em>consolidation</em> (=2) and <em>pleural effusion</em> (=3). We then trained a 2d multilabel U-Net model, which you can find and apply in MedSeg

本数据集包含来自40余名新冠患者的100张轴向CT图像，这些图像由公开可获取的JPG格式图像转换而来，原始图像来源网址为https://www.sirm.org/en/。其图像转换流程已在博文《Covid-19 radiology — data collection and preparation for Artificial Intelligence》中详细说明。简言之，本数据集的图像已由放射科医师完成分割标注，共设置3类标签：磨玻璃影（ground-glass，掩码值=1）、实变影（consolidation，掩码值=2）与胸腔积液（pleural effusion，掩码值=3）。随后我们基于该数据集训练了二维多标签U-Net模型，该模型可在MedSeg平台获取并使用。

搜集汇总

背景与挑战

背景概述

该数据集包含100张COVID-19患者的轴向CT图像，经过放射科医生标注，并提供了用于医学图像分割的预训练模型。