1.2.3.2 编码孔径断层合成

1.2.3 Coded Aperture Tomosynthesis

One of the most intriguing and advanced tomosynthesis methods of the past, coded aperture imaging, was developed by Klotz and Weiss of Philips GmbH of Hamburg, Germany in the 1970s (Klotz and Weiss 1974, 1976, Weiss et al. 1977, 1979). This method could generate arbitrary tomosynthesis planes with imaging times of only milliseconds, essentially eliminating any problems associated with patient motion. They called their method “flashing tomosynthesis,” and its application to coronary angiography was described in a paper that they coauthored with Woelke et al. (1982). The coding and decoding steps are illustrated in Figure 1.7. In general, for coding, many x-ray sources were either pulsed sequentially or simultaneously (Figure 1.7a). A single 60 × 60 cm film recorded the entire set of subimages, forming the coded image. For decoding, this coded image was illuminated with a light box and a three-dimensional (3D) image of the object in space was produced with an array of lenses that were arranged according to the distribution of x-ray tubes. A ground glass screen (Figure 1.7b) was positioned within the 3D image to display different layers within the object.

Figure 1.7 coding step (a) and decoding step (b) of the flashing (coded aperture) tomosynthesis system described in Woelke et al. (reproduced with permission from Radiology.)
The system Woelke et al. employed consisted of 24 small stationary x-ray tubes that were fired simultaneously. The exposure time was about 50 ms, the radiation exposure to the skin was about 1 roentgen, and the slice thickness was about 1 mm. In addition to the multiple x-ray tubes, the hardware in the system included an optical postprocessing unit, a TV monitor on which reconstructed layers were displayed, and a film hardcopy unit (Woelke et al. 1982). In a study of 20 left coronary artery stenoses within 10 postmortem hearts that were placed within a thorax phantom, Woelke et al. found that the correlation between the degree of stenosis determined with flashing tomosynthesis and morphometry (r = 0.92, p < 0.001, SEE = 9%) was better than the correlation between the degree of stenosis determined with conventional 35-mm cine film cardiac imaging and morphometry (r = 0.82, p < 0.001, SEE = 16%). They also successfully employed their flashing tomosynthesis technique on five patients with coronary artery disease. Limitations of the flashing tomosynthesis method included restriction to a small 9-cm-diameter field of view, and the inability to evaluate blood flow dynamics due to the use of a single “flash” exposure (Woelke et al. 1982). Additional papers that have been published on flashing and coded aperture tomosynthesis include those of Groh (Groh 1977), Nadjmi et al. (1980), Stiel et al. (Stiel 1989, 1992, 1993), Haaker et al. (1985a,b, 1990), and Becher et al. (1983, 1985).

1.2.3.2 编码孔径断层合成

编码孔径成像，是过去最有趣、最先进的断层合成方法之一，它是由德国汉堡的飞利浦公司的 Klotz和Weiss在1970年代开发的（ Klotz and Weiss 1974, 1976, Weiss et al. 1977, 1979）。这种方法可以任意的断层合成平面，成像时间仅为毫秒，从而基本上消除了与患者运动有关所导致的问题。他们称之为“山火断层合成”，并在他们与Woelke等人合著的论文中描述了其在冠状动脉造影中的应用（1982）。编码和解码的步骤如图1.7所示。通常，进行编码时，许多X射线源同时或顺序的发生脉冲（图1.7a）。一张60x60厘米的胶片记录下了整个子图像集，形成了编码图像。进行解码时，用灯箱对编码图像进行照明，并使用依据X射线管分布排列的透镜阵列生成空间中物体的三维图像。将毛玻璃屏幕（图1.7b）放置在3D图像内，以显示对象内的不同层。

图1.7 Woelke等人描述的闪烁（编码孔径）断层合成系统中的编码和解码步骤。
Woelke所使用的系统包括超过24个同事发射的小型固定X射线管。曝光时间约为50ms，暴露于皮肤的辐射约为1伦琴，切片厚度为1mm。除了多个X射线管之外，系统中的硬件还包括光学后处理单元，电视监视器用于显示重构层以及胶片硬拷贝单元（Woelke等，1982）。 Woelke等人在一项研究中对放置在胸模内的10个死后的心脏中的20个左冠状动脉狭窄进行了研究，发现通过闪烁断层合成和形态学测量确定的狭窄程度的相关性（r = 0.92，p <0.001，SEE = 9％）优于传统的35毫米电影胶片心脏成像和形态测量确定的狭窄程度之间的相关性（r = 0.82，p <0.001，SEE = 16％）。他们还成功地将闪闪烁断层合成技术应用于5例冠心病患者。闪烁断层合成的局限性包括视场限制在9cm以内，以及由于仅使用了一次曝光而无法评估血流动力学（Woelke等，1982）。关于闪烁和编码孔径断层合成的其他论文包括Groh（1977），Nadjmi （1980），Stiel（1989，1992，1993），Haaker （1985a，b，1990），和Becher等 （1983，1985）。