Penn Image Computing and Science Laboratory, University of Pennsylvania School of Medicine, 3600 Market Street, Suite 370, Philadelphia, PA 19104, USA.
Acad Radiol. 2012 Jul;19(7):857-64. doi: 10.1016/j.acra.2012.03.007. Epub 2012 Apr 18.
The aim of this study was to compare the performance of various image-based metrics computed from thoracic high-resolution computed tomography (HRCT) with data from pulmonary function testing (PFT) in characterizing interstitial lung disease (ILD) and chronic obstructive pulmonary disease (COPD).
Fourteen patients with ILD and 11 with COPD had undergone both PFT and HRCT within 3 days. For each patient, 93 image-based metrics were computed, and their relationships with the 21 clinically used PFT parameters were analyzed using a minimal-redundancy-maximal-relevance statistical framework. The first 20 features were selected among the total of 114 mixed image metrics and PFT values in the characterization of ILD and COPD.
Among the best-performing 20 features, 14 were image metrics, derived from attenuation histograms and texture descriptions. The highest relevance value computed from PFT parameters was 0.47, and the highest from image metrics was 0.52, given the theoretical bound as [0, 0.69]. The ILD or COPD classifier using the first four features achieved a 1.92% error rate.
Some image metrics are not only as good discriminators as PFT for the characterization of ILD and COPD but are also not redundant when PFT values are provided. Image metrics of attenuation histogram statistics and texture descriptions may be valuable for further investigation in computer-assisted diagnosis.
本研究旨在比较从胸部高分辨率 CT（HRCT）计算的各种基于图像的指标与肺功能测试（PFT）数据在特征化间质性肺病（ILD）和慢性阻塞性肺疾病（COPD）中的表现。
14 名患有ILD 的患者和 11 名患有 COPD 的患者在 3 天内同时接受了 PFT 和 HRCT。对于每个患者，计算了 93 个基于图像的指标，并使用最小冗余最大相关性统计框架分析了它们与 21 个临床使用的 PFT 参数之间的关系。在ILD 和 COPD 的特征描述中，从总共 114 个混合图像指标和 PFT 值中选择了前 20 个特征。
在表现最佳的 20 个特征中，有 14 个是来自衰减直方图和纹理描述的图像指标。从 PFT 参数计算出的最高相关性值为 0.47，而从图像指标计算出的最高相关性值为 0.52，理论上限为[0,0.69]。使用前四个特征的ILD 或 COPD 分类器的错误率为 1.92%。
一些图像指标不仅与 PFT 一样可以很好地区分ILD 和 COPD 的特征，而且在提供 PFT 值时也不具有冗余性。衰减直方图统计和纹理描述的图像指标可能对计算机辅助诊断的进一步研究具有重要价值。
