Fenske R A, Birnbaum S G
Department of Environmental Health, School of Public Health and Community Medicine, University of Washington, Seattle 98195, USA.
Am Ind Hyg Assoc J. 1997 Sep;58(9):636-45. doi: 10.1080/15428119791012423.
Development of a second-generation video imaging technique for assessing occupational skin exposure (VITAE) is described, its performance evaluated, and new procedures for exposure quantification are presented. The current VITAE system has higher resolution in regard to both its picture element array and gray scale when compared with the prototype system. System performance was evaluated during extended field deployment: variability was 3-4% during data acquisition for individual worker evaluation session, and 10% over a 22-day study period. Variabilities attributable to subject positioning and image outlining procedures were 2.7 and 1.2%, respectively. Visual observations of fluorescent tracer deposition on skin were used to classify specific body regions as either exposed of unexposed, and two computer-based classification criteria were tested against the visual classification. These criteria were generally better at minimizing false negative and false positive classification; sensitivity and predictive value reached 95 and 99%, respectively, when analysis was preceded by presampling of a subset of images. Variability in skin pigmentation was found to have a substantial effect on fluorescent tracer qualification, leading to development of new calibration procedures. Standard curves were generated by spotting a range of tracer concentrations on volunteer subjects and quantifying fluorescence with the VITAE system. These data were then grouped either by subject or by the magnitude of the background signal of the unexposed skin. The ability to control for the effects of skin pigmentation was found to be comparable for these grouping methods, indicating that calibration curves can be developed without the creation of a unique curve for each subject.
本文描述了用于评估职业性皮肤暴露的第二代视频成像技术(VITAE)的开发过程,评估了其性能,并介绍了暴露量量化的新程序。与原型系统相比,当前的VITAE系统在像素阵列和灰度方面具有更高的分辨率。在扩展的现场部署期间对系统性能进行了评估:在单个工人评估期间的数据采集过程中,变异性为3-4%,在为期22天的研究期间为10%。受试者定位和图像轮廓绘制程序导致的变异性分别为2.7%和1.2%。通过对皮肤上荧光示踪剂沉积的视觉观察,将特定身体区域分类为暴露或未暴露,并针对视觉分类测试了两种基于计算机的分类标准。这些标准通常在最小化假阴性和假阳性分类方面表现更好;在对一部分图像进行预采样后进行分析时,灵敏度和预测值分别达到95%和99%。发现皮肤色素沉着的变异性对荧光示踪剂的定量有重大影响,从而导致了新校准程序的开发。通过在志愿者受试者身上滴加一系列示踪剂浓度并用VITAE系统对荧光进行定量,生成了标准曲线。然后,这些数据按受试者或未暴露皮肤背景信号的大小进行分组。发现这两种分组方法在控制皮肤色素沉着影响方面的能力相当,这表明可以在不针对每个受试者创建唯一曲线的情况下开发校准曲线。