CSI 2014 Vertebrae Localization and Identification

脊柱CT成像中椎骨的准确定位和识别对于许多临床任务 (例如诊断，手术计划和术后评估) 很重要。临床数据集给自动方法带来了许多困难。这些原因是由于外科植入物引起的脊柱弯曲异常，视野小以及图像伪影的频繁存在。为了促进该主题的研究进展，我们提供了242注释的脊柱ct扫描的数据库。这些扫描可用于开发，培训和评估。此外，现在为挑战的测试阶段提供了60个数据集，因此不会提供任何手动注释。测试数据的结果需要提交到在线评估系统。Accurate localization and identification of vertebrae in spinal computed tomography (CT) images is critical for a wide range of clinical tasks, including diagnosis, surgical planning, and postoperative assessment. Clinical spinal CT datasets present significant challenges to automated analysis methods. These challenges stem from abnormal spinal curvature induced by surgical implants, restricted field of view, and the frequent occurrence of image artifacts. To facilitate research progress in this field, we release a database comprising 242 annotated spinal CT scans, which can be employed for developing, training, and evaluating automated algorithms. Furthermore, 60 datasets are provided for the challenge's test stage, with no manual annotations disclosed. Results obtained from the test datasets must be submitted to the online evaluation system.

搜集汇总

背景与挑战

背景概述

该数据集是一个公开的脊柱CT成像数据集，专注于椎骨的准确定位和识别，旨在支持临床任务如诊断和手术计划。它包含242个已注释的CT扫描用于训练和开发，以及60个未注释的测试数据用于在线评估挑战，由格拉茨理工大学在2014年发布，适用于智慧医疗领域的研究。