Dataset (features extracted from chest CT images) accompanying the paper Automatic emphysema detection using weakly labeled HRCT lung images

The dataset contains derived features from CT images of patients and controls scanned at two different centers, Frederikshavn and Aalborg.Each image is represented by 50 feature vectors, where each feature vector describes a volumetric ROIs of size 41 x 41x 41 voxels, extracted at random locations inside the lung mask. The features extracted are Gaussian scale space features, or histograms of intensity values in the ROI after filtering the image. Here we use eight filters (smoothed image, gradient magnitude, Laplacian of Gaussian, three eigenvalues of the Hessian, Gaussian curvature and eigen magnitude), four scales (0.6, 1.2, 2.4 and 4.8 mm), and histograms of ten bins. <br>

本数据集包含来自两家不同扫描中心——腓特烈港（Frederikshavn）与奥尔堡（Aalborg）——的患者及健康对照者的CT影像衍生特征。每份影像对应50个特征向量，每个特征向量对应一个尺寸为41×41×41体素的三维感兴趣区域（Region of Interest, ROI），该区域从肺掩码内的随机位置提取。所提取的特征为高斯尺度空间特征，或是对影像完成滤波后，感兴趣区域内的强度值直方图。本研究使用了八种滤波器（平滑图像、梯度幅值、高斯拉普拉斯、海森矩阵（Hessian）的三个特征值、高斯曲率以及特征幅值）、四种尺度参数（0.6、1.2、2.4与4.8毫米），以及含10个分箱的直方图。