【作者】 李冰；

【导师】 王玉瑾；

【作者基本信息】 山西医科大学 ， 法医学， 2014， 博士

【摘要】 应用口腔放射影像资料进行同一认定,是目前法医牙科学一项重要的研究课题。口腔颌面X线曲面断层片或X线头颅侧位片同一认定特征指标的选择,是根据口腔颌面部的牙齿、颅骨、颌骨、腭皱及其毗邻结构的具有独特形态特征的生理解剖特征、病理变化和牙齿的各种治疗情况等进行的。根据口腔颌面x线片同一认定的方法与指标还有待进一步客观化、标准化。近年来随着口腔保健水平的提高以及牙科治疗手段的不断进步,牙齿发生病变的情况在逐渐减少。单纯依靠牙齿病变和治疗特征进行同一认定已不能完全适应新的情况。有必要探寻新的识别指标来解决无牙科治疗者的法医学同一认定。口腔颌面部软组织腭皱的形态,是由遗传基因控制的个体特征,在人的一生中从不改变,甚至疾病,外伤或化学腐蚀都不会改变腭皱的形态和位置。因为人类腭皱形态的独特性及稳定性,使其逐渐成为法医同一认定中又一新的的识别标志。本研究分别对随机抽取的200张牙齿有治疗史者的X线曲面断层片及600张X线曲面断层片(牙齿或治疗或没有治疗)的全口牙列不同区域进行全面观测,根据牙齿排列特征及牙齿病变和治疗特征进行初级编码和分级编码,统计其牙列编码的多样性。本研究还选取200张x线头颅侧位片,通过SNA角、SNB角的相互关系以及后面高(S-Go)与前面高(N-Me)比值,对颌面骨骼形态进行矢状向8类分型及垂直向3类的综合分类,同时筛选出8个测量指标对颌面特征进行数字编码,为利用口腔X线头影测量片进行同一认定提供理论依据。此外,本课题结合腭皱的形状、数量、位置、分布等特征,对200例腭皱的形态图进行全面系统的编码、统计分析及评估,以期建立腭皱形态学识别系统的同一认定指标。同时,在标准条件下采集腭皱图像信息,通过腭皱数字图像识别系统,进行图像降噪,图像增强,图像分割,特征提取,边缘检测,信息匹配,结合Matlab软件处理,实现更快更准确的腭皱图像辨别。并期望对腭皱形态的收集,记录和计算机分析提出标准和统一的程序,为法医同一认定中提供一种有效方法。研究结果显示：(1)600张x线曲面断层片(牙齿或治疗或没有治疗)中,共有544种类型的全口牙列编码,全口牙列编码多样性为90.67%,其中上颌牙列编码的多样性为74.84%,下颌牙列编码的多样性为74.33%。200张牙齿有治疗史者的X线曲面断层片中,共有198种类型的全口牙列编码,全口牙列编码多样性为99.0%,其中上颌牙列编码的多样性为98.0%,下颌牙列编码的多样性为98.5%。(2)600张X线曲面断层片(牙齿或治疗或没有治疗)中,前牙区牙列编码的多样性仅为24.33%,而牙齿有治疗史者的前牙区牙列编码的多样性高于随机抽取组为59.50%,两组各个后牙区牙列编码的多样性均大于前牙。(3)提出X线头颅侧位片矢状向9类及垂直向3类共27类面型分型,筛选出牙颌面8项定量分级识别指标变异系数较大,一些分级测量指标性别差异有统计学意义(P<0.05)。(4)200例受试者腭皱形态编码的多样性为100%,个体间未见腭皱类型完全一致,每个个体不论男性与女性均表现有独特的腭皱形式,且同一个体左右侧腭皱形态及分布亦不同。其中波浪形腭皱形态所占比例最大(23.03%),三分叉形态出现比例最小(0.74%)。不同性别腭皱形态所占比例,波浪形及曲线形腭皱形态所占比例均较大,女性：波浪形(22.7%)及曲线形(18.28%)；男性：波浪形(24.11%)及曲线形(21.43%)。(5)对腭皱进行图像采集、特征提取、边缘检测、信息匹配,结合Matlab软件处理,能实现更快更准确的腭皱图像辨别。本研究提示,有牙齿治疗特征者全口牙列编码、后牙区牙列编码的方法对于这类人群的法医学同一认定有非常高的价值；采用X线头颅侧位片进行同一认定时可先作矢状向及垂直向分型鉴别,再联合用所选取的8项定量分级识别指标,对牙颌面形态进行识别编码,口腔颌面部相关组织软硬组织具有独特形态特征的解剖结构,在法医学鉴定中可用于同一认定。更多还原

【Abstract】 Identification using oral radiographic image materials is one of the important study directions in forensic dentistry medicine The characteristic index of maxillofacial for forensic identification were chosen according to anatomical structures,pathological changes and dental interventions of teeth and maxillofacial bone by the X-ray orthopantomography or X-ray lateral cephalogram.However,the methods and items of personal identification using radiographic image materials should be more objectified and standardized.But in recent years, the progress of dental health and treatment have resulted in a gradually decrease in dental caries. As a result of that the forensic identification only by characteristics dental caries and interventions cannot fully adapt to the new situation. Therefore, it is necessary to explore new index for identifying the free dental treatment to solve the forensic identification.Palatal rugae is mediated by genetic genes, showing various characteristics among population and retains unchanged in each individual.In the human lifetime,the length of palatal rugae may change but the location maintains unchanged,it never alters by disease, trauma or chemical corrosion.Due to unique and diverse features, human palatal rugae attract increasing attention as a new marker in the forensic identification.All of the the digital X-ray orthopantomogram were comprehensively observed, including200samples with dental interventions,600samples(treatmen or treatmenless) randomly selected.in this study indexes were selected for dentition patterns according to the characters of dental alignment,dental pathological changes and intervention patterns.Two hundreds X-ray lateral cephalogram were selected.Facial pattern were classified by sagittal and vertical features. Maxillofacial sagittal features is divided into9categories in accordance with the relationship between SNA and SNB, According to the ratio of S-Go and N-Me,vertical facial pattern were classified into3categories.Eight identification indexes were identified to code the identification indexes for craniofacial morphology.It will provide a theoretical basis for the use of lateral cephalogram in Forensic Identification Indexes. The figures of palatal rugae of200study models were depicted. The classification was followed based on form、quantity、position and distribution to assess the individuality of rugae pattern,and expected to construct a forensic identification system for oral palatal rugae.we simultaneously collected the information of rugae images under standard conditions, and used Rugae Image Identification System (RIIS) in various ways including collection, noise reduction, augmentation, segmentation, feature extraction, edge detection, and information matching, with each a corresponding algorithm. Together with the processing of Matlab, we succeeded in obtaining a more rapid and accurate way of rugae image identification. We also suggested a standard and unified process for the collection, documentation, and computer analysis of rugae patterns, hoping this will offer an effective method for forensic identification. The findings:(1) The group with randomly selected with or without therapy had544kinds of dental pattern in600samples,thus its diversity of full dentition patternswas90.67%.The diversity was74.84%in maxillary and74.33%in mandible.The group with dental interventions had198kinds of dental pattern in200samples,thus its diversity of full dentition patterns was99.0%.The diversity was98.0%in maxillary and98.5%in mandible.(2) In600samples,the diversity in anterior regions is24.33%,which was lower than the samples with dental interventions(59.50%).Both of these groups showed a tendency of the diversity in posterior regions was higher than anterior regions.(3) Facial pattern were classified into27categories.According to9sagittal and3vertical facial pattern.All eight quantitative variables had comparatively high CV value.The difference of classification index between both sexes has statistical significance(p<0.05).(4) Among the involved200models, palatal rugae types varied among individuals,the diversity of palatal rugae types is100%.Palatal rugae pattern was different between men and women, even between two sides in the same individual. The curve type accounted for the higher percentage (23.03%), while the trifurcated type occupied the lowest proportion (0.74%).The sinuous and curve types were dominant in men and women (women:sinuous22.7%, curve18.28%; men:sinuous24.11%, curve21.43%)(5) Image Identification in various ways including collection, noise reduction, augmentation, segmentation, feature extraction, edge detection, and information matching, with each a corresponding algorithm. Together with the processing of Matlab, we succeeded in obtaining a more rapid and accurate way of rugae image identification.Base on these findings, it is concluded that the group with dental interventions had very good full dentitiondiversity in full dentition involved the posterior regions.Thus the method of full or part dentition coding was valuable in the group. Initial identification by the Facial pattern classification of Maxillofacial sagittal features followed with coding the craniofacial morphology with8indexes is likely a novel approach for forensic identification of in X-ray lateral cephalogram.the study shows the characteristic index of maxillofacial for forensic identification were chosen according to soft-sclerous tissues morphologic characteristic更多还原