Henze E, Graf G, Clausen M, Rall B, Weller R, Derichs D, Kreidler J, Heidenreich P, Sitzmann F, Adam W E
Division of Nuclear Medicine, University of Ulm, Federal Republic of Germany.
Eur J Nucl Med. 1990;16(2):97-101. doi: 10.1007/BF01465918.
The exact regional correlation of findings of facial bone scans, planar or SPECT, to dental orthopan X-ray films (OPT) is difficult because of the very different projection techniques. To improve correlative imaging in this regard a projection algorithm was developed that uses SPECT data of the skull for reconstructing an orthopan tomoscintigraphic projection. Fourteen conventional SPECT slices of the upper and lower jaws were obtained during bone scanning. All mandibular slices were superimposed resulting in a horseshoe shaped structure, while was marked by an ROI which was divided into segments. All 14 SPECT slices were then masked by this segmental ROI, thereby marking the teeth-carrying bone in all slices. The information from this horseshoe like ROI is then transformed into lines. Line by line arrangement results in an orthopan projection, the orthopan tomoscintigram. This new display allows 1:1 true scale superimposition with the X-ray OPT and markedly facilitates correlative imaging.
由于投影技术差异很大,面部骨扫描(平面或SPECT)结果与牙科全景X线片(OPT)的确切区域相关性很难确定。为了在这方面改善相关成像,开发了一种投影算法,该算法使用颅骨的SPECT数据来重建全景断层闪烁投影。在骨扫描过程中获得了上下颌的14个常规SPECT切片。所有下颌切片叠加在一起形成一个马蹄形结构,该结构由一个被划分为多个节段的感兴趣区(ROI)标记。然后用这个节段性ROI对所有14个SPECT切片进行掩膜,从而在所有切片中标记出含牙骨。然后将来自这个马蹄形ROI的信息转换为线条。逐行排列形成一个全景投影,即全景断层闪烁图。这种新的显示方式允许与X线OPT进行1:1的真实比例叠加,并显著促进相关成像。