multivendor dataset

本研究使用了一个名为multivendor dataset的数据集，包含来自9个中心和6个不同供应商的613个数字乳腺摄影图像。数据集主要用于比较视觉评估与全自动计算的乳腺密度。数据集中的图像涵盖了多种设备，包括Amulet FDR 1000、Senograph DS、Selenia Dimensions等。数据集的创建旨在解决乳腺密度评估的主观性和可重复性问题，特别是在乳腺摄影的临床实践中。通过此数据集，研究者能够评估自动化乳腺密度评估器（ABDE）与经验丰富的乳腺放射科医生之间的评估一致性，从而为个性化筛查策略提供依据。

This study utilized a dataset named the multivendor dataset, which comprises 613 digital mammography images sourced from 9 institutions and 6 distinct vendors. This dataset is primarily intended to compare visual assessment of breast density with fully automated computed breast density. The images included in the dataset cover a diverse range of equipment models, including Amulet FDR 1000, Senograph DS, Selenia Dimensions, and other similar devices. The dataset was developed to address the problems of subjectivity and poor reproducibility in breast density assessment, particularly within the clinical practice of mammography. Leveraging this dataset, researchers can evaluate the consistency of breast density assessment between automated breast density evaluators (ABDE) and experienced breast radiologists, thereby providing evidence to support personalized screening strategies.