Meintjes E M, Douglas T S, Martinez F, Vaughan C L, Adams L P, Stekhoven A, Viljoen D
MRC/UCT Medical Imaging Research Unit, Department of Human Biology, Faculty of Health Sciences, University of Cape Town, 7925, Observatory, South Africa.
Med Eng Phys. 2002 Dec;24(10):683-9. doi: 10.1016/s1350-4533(02)00114-5.
In diagnosing a child with fetal alcohol syndrome (FAS), anthropometric measurements of the face are conventionally performed by highly trained dysmorphologists using a hand-held ruler. This renders the screening of large populations of children for the facial features characteristic of FAS very time-consuming and costly. This study proposes a new, cost-effective, and non-intrusive method to measure in three dimensions the facial dysmorphology of children using stereo-photogrammetry. The face of each child is photographed in a control frame simultaneously by a pair of high-resolution digital cameras mounted 1.04 m from the child and 0.26 m apart. Software has been developed to calibrate the images and to compute the three-dimensional object-space coordinates of any point on the face from a measurement of the point on each of the images. The palpebral fissure lengths, inner canthal-, and interpupillary distances of 44 subjects were measured in this manner independently by two investigators and compared with measurements obtained by clinical specialists in the conventional manner. There was found to be no statistically significant difference between palpebral fissure lengths determined using the two techniques (paired Student's t-test p-values are 0.29 and 0.18, respectively). It has also been demonstrated that facial measurements can be performed with greater consistency from a pair of stereo photographs than direct measurements from live subjects.
在诊断患有胎儿酒精综合征(FAS)的儿童时,面部人体测量传统上由训练有素的畸形学家使用手持直尺进行。这使得对大量儿童进行FAS特征性面部特征筛查既耗时又昂贵。本研究提出了一种新的、具有成本效益且非侵入性的方法,使用立体摄影测量法对儿童面部畸形进行三维测量。每个儿童的面部由安装在距离儿童1.04米、相距0.26米的一对高分辨率数码相机在控制框架内同时拍摄。已开发出软件来校准图像,并根据对每张图像上某点的测量计算出该点在面部的三维物体空间坐标。44名受试者的睑裂长度、内眦间距和瞳孔间距由两名研究人员以这种方式独立测量,并与临床专家以传统方式获得的测量结果进行比较。发现使用这两种技术确定的睑裂长度之间没有统计学上的显著差异(配对学生t检验p值分别为0.29和0.18)。还证明了从一对立体照片进行面部测量比从活体受试者直接测量具有更高的一致性。