【摘要】 本文提出了一种能够半自动地从医学CT图像中获取颅骨轮廓标志点的方法,快速准确地从54组儿童头部CT扫描数据中提取标志点,并通过主成分分析和回归分析,建立了儿童颅骨轮廓点集与儿童宏观参数(年龄和头部周长)之间的回归关系模型,并据之得出了6个月、18个月、36个月儿童头部轮廓形态的统计结果。结果表明,该模型可较好地预测给定宏观参数的儿童颅骨轮廓。更多还原

【Abstract】 A technique is proposed to semi-automatically acquire the marking points of skull contour from medical CT images,and with which marking points are quickly and accurately extracted from 54 sets of CT scanned data of children heads. Then by using principal component analysis and regression analysis,a regression relationship model between the skull contour point set and the macro-parameters of a child( e. g. age and head circumference) is established,and based on which the statistical results of 6-,18-,and 36- month-old children head skull contours are obtained. It is shown that the model established can better predict the skull contours of children with given macro-parameters.更多还原