Department of Biomedical Engineering, Case Western Reserve University, 319 Wickenden Building, 10900 Euclid Avenue, Cleveland, Ohio 44106, USA.
Med Phys. 2011 Jan;38(1):107-13. doi: 10.1118/1.3523098.
To determine the diagnostic efficacy of optical coherence tomography (OCT) to identify cervical intraepithelial neoplasia (CIN) grade 2 or higher by computer-aided diagnosis (CADx).
OCT has been investigated as a screening/diagnostic tool in the management of preinvasive and early invasive cancers of the uterine cervix. In this study, an automated algorithm was developed to extract OCT image features and identify CIN 2 or higher. First, the cervical epithelium was detected by a combined watershed and active contour method. Second, four features were calculated: The thickness of the epithelium and its standard deviation and the contrast between the epithelium and the stroma and its standard deviation. Finally, linear discriminant analysis was applied to classify images into two categories: Normal/inflammation/CIN 1 and CIN 2/CIN 3. The algorithm was applied to 152 images (74 patients) obtained from an international study.
The numbers of normal/inflammatory/CIN 1/CIN 2/CIN 3 images are 74, 29, 14, 24, and 11, respectively. Tenfold cross-validation predicted the algorithm achieved a sensitivity of 51% (95% CI: 36%-67%) and a specificity of 92% (95% CI: 86%-96%) with an empirical two-category prior probability estimated from the data set. Receiver operating characteristic analysis yielded an area under the curve of 0.86.
The diagnostic efficacy of CADx in OCT imaging to differentiate high-grade CIN from normal/low grade CIN is demonstrated. The high specificity of OCT with CADx suggests further investigation as an effective secondary screening tool when combined with a highly sensitive primary screening tool.
通过计算机辅助诊断（CADx）确定光学相干断层扫描（OCT）识别宫颈上皮内瘤变（CIN）2 级及以上的诊断效能。
OCT 已被研究作为一种筛查/诊断工具，用于管理宫颈的早期癌前病变和早期浸润性病变。在本研究中，开发了一种自动算法来提取 OCT 图像特征并识别 CIN 2 级及以上病变。首先，采用分水岭和主动轮廓相结合的方法检测宫颈上皮。其次，计算了四个特征：上皮厚度及其标准差，上皮与基质之间的对比度及其标准差。最后，应用线性判别分析将图像分为两类：正常/炎症/CIN1 和 CIN2/CIN3。该算法应用于 152 张图像（74 例患者），这些图像来自一项国际研究。
正常/炎症/CIN1/CIN2/CIN3 图像的数量分别为 74、29、14、24 和 11。十折交叉验证预测该算法的敏感性为 51%（95%CI：36%-67%），特异性为 92%（95%CI：86%-96%），其经验两分类先验概率是从数据集估计的。受试者工作特征曲线分析得出曲线下面积为 0.86。
CADx 在 OCT 成像中区分高级别 CIN 与正常/低级别 CIN 的诊断效能得到了验证。OCT 结合 CADx 的高特异性表明，当与高敏感性的初级筛查工具结合使用时，可进一步将其作为一种有效的二级筛查工具进行研究。
