Thekkek Nadhi, Lee Michelle H, Polydorides Alexandros D, Rosen Daniel G, Anandasabapathy Sharmila, Richards-Kortum Rebecca
Rice University, Department of Bioengineering, MS-142, Box 1892, Houston, Texas 77251-1892, United States.
Icahn School of Medicine, Mount Sinai Medical Center, One Gustave L. Levy Place, Box 1069, New York, New York 10029-6574, United States.
J Biomed Opt. 2015 May;20(5):56002. doi: 10.1117/1.JBO.20.5.056002.
Current imaging tools are associated with inconsistent sensitivity and specificity for detection of Barrett's-associated neoplasia. Optical imaging has shown promise in improving the classification of neoplasia in vivo. The goal of this pilot study was to evaluate whether in vivo vital dye fluorescence imaging (VFI) has the potential to improve the accuracy of early-detection of Barrett's-associated neoplasia. In vivo endoscopic VFI images were collected from 65 sites in 14 patients with confirmed Barrett's esophagus (BE), dysplasia, oresophageal adenocarcinoma using a modular video endoscope and a high-resolution microendoscope(HRME). Qualitative image features were compared to histology; VFI and HRME images show changes in glandular structure associated with neoplastic progression. Quantitative image features in VFI images were identified for objective image classification of metaplasia and neoplasia, and a diagnostic algorithm was developed using leave-one-out cross validation. Three image features extracted from VFI images were used to classify tissue as neoplastic or not with a sensitivity of 87.8% and a specificity of 77.6% (AUC = 0.878). A multimodal approach incorporating VFI and HRME imaging can delineate epithelial changes present in Barrett's-associated neoplasia. Quantitative analysis of VFI images may provide a means for objective interpretation of BE during surveillance.
当前的成像工具在检测巴雷特食管相关肿瘤时,其敏感性和特异性存在不一致的情况。光学成像在改善体内肿瘤的分类方面已显示出前景。这项初步研究的目的是评估体内活体染料荧光成像(VFI)是否有潜力提高巴雷特食管相关肿瘤早期检测的准确性。使用模块化视频内窥镜和高分辨率微型内窥镜(HRME),从14例确诊为巴雷特食管(BE)、发育异常或食管腺癌的患者的65个部位收集了体内内窥镜VFI图像。将定性图像特征与组织学进行比较;VFI和HRME图像显示了与肿瘤进展相关的腺结构变化。确定了VFI图像中的定量图像特征,用于化生和肿瘤的客观图像分类,并使用留一法交叉验证开发了一种诊断算法。从VFI图像中提取的三个图像特征用于将组织分类为肿瘤性或非肿瘤性,敏感性为87.8%,特异性为77.6%(AUC = 0.878)。结合VFI和HRME成像的多模态方法可以描绘出巴雷特食管相关肿瘤中存在的上皮变化。VFI图像的定量分析可能为监测期间BE的客观解读提供一种方法。