利用第二頸椎之電腦斷層資料推估台灣人身高

Abstract

死者身高推估為法醫人類學中重要研究領域，其所得的資訊可以提供人身鑑別更多線索，縮小死者身份調查範圍。過往研究中，主要以長骨為研究對象，但在近年來大災難等事件中，常見碎屍塊與骨骸殘片而缺乏長骨。本研究目的為藉由易於辨識的第二頸椎結合電腦斷層影像技術進行身高推估公式研究，以期未來在骨骸殘片遺體推估身高時能有國人合適之推估公式應用。 本研究採用回溯性研究，收集2010年1月至2016年12月止之臺大醫院影像醫學部頭頸部電腦斷層資料。研究樣本為389人，電腦斷層影像資料由影像處理評估軟體RadiAnt DICOM Viewer （64-bit）讀取，量測第二頸椎之齒突頂點至椎體前端最低點位（the top of the dens to the anterior-inferior point of the vertebral body，縮寫：DA）、錐體前端最低點至棘突頂點（the anterior-inferior point of the vertebral body to the posterior point of the spinous process，縮寫：AS）與棘突頂點至齒突頂點（the top of the dens to the posterior point of the spinous process，縮寫：DS）之最長長度。在Intra-rater與Inter-rater重複性試驗在Cronbach's alpha檢視信度下，其alpha值皆≧0.9，其測量結果一致性表現佳。 研究結果共分為全部389人組、男性251人組、女性138人組共三組，每組皆得到單一特徵點位長度、混合兩種與三種特徵點位長度組合身高推估計算公式共六組公式，總計18組身高推估公式。並以研究樣本389人的15％為驗證validation人數，另調閱59人（性別分佈為男性37人，女性22人）頭頸部電腦斷層影像資料，運用Paired t test進行方法效度分析檢定，以檢驗本研究所統計出身高推論公式方法效度。除女性138人組中DS單一特徵點位身高推估公式、DA + AS兩項特徵點位身高推估公式、DS + DA兩項特徵點位身高推估公式、DA + AS + DS三項特徵點位身高推估公式以上四組身高推估公式，其效度檢定結果不適用外，其餘在不分性別組別與男性組別推估身高皆與原始身高無顯著差異。 本研究建立之臺灣人第二頸椎身高推估公式，在法醫實務上不分性別的遺體或骨骸屍等狀況，最適合採用之推估公式為AS（錐體前端最低點至棘突頂點）單一特徵點位身高推估公式，其效度分析結果最佳。

Human body height prediction from residual bones is an important field in forensic anthropology. It may provide information in individual identification and can apply to judicial investigation. In previous studies, researchers focused on the association between long bones and body height. However, we may obtain only fragments of long bones and small bones for body height prediction in mass disaster. This study aimed to develop formulas to estimate adult body height of different genders by using computed tomography data of human second cervical vertebrae of Taiwanese. The results may apply to body height prediction with residual bones of anonymous corpses in forensic casework. This research was a retrospective study. Our samples were obtained from the Department of Radiology National Taiwan University Hospital from January 2010 to December 2016. The sample number of head and neck computed tomography was 389, including 251 males and 138 females (age range was 20-80 years old; mean number was 51.5 years old). The medial sagittal view images of second vertebrae were measured by RadiAnt DICOM Viewer (64-bit) software. The measurements were taken at three points and three lengths were obtained: the top of the dens to the anterior-inferior point of the vertebral body (DA); the anterior-inferior point of the vertebral body to the posterior point of the spinous process (AS); the top of the dens to the posterior point of the spinous process (DS). In this study, 18 human height prediction formulas from second cervical vertebrae was established. The reliability analysis in intra-rater and inter-rater repeatability showed good consistency, with all the Cronbach’s alpha above 0.9. The paired-t test was used for validation analysis. There were no significant differences between the observed body heights of these subjects and the estimated heights of subjects in both-gender group and male group. However, in the female group, there were four formulas failed for body height prediction. In conclusion, we contrasted the formulas of Taiwanese body height from second cervical vertebrae by using computed tomography data. We recommend the use of AS formula in body height prediction for unknown-gender corpuses in forensic casework.