Research On Forensic Individual Identification Based On 3D Model Of Frontal Sinus Using Multi-slice Spiral Computed Tomography

Objective Based on the computerized tomography image,the three-dimensional model of the frontal sinus is established,and the observation index is selected to classify the three-dimensional morphological features of the frontal sinus.At the same time,the coding study is combined with the digital coding of the index,which provides a theoretical basis for forensic individual identification using the frontal sinusMethods The plain CT images of 476 patients aged from 21 to 70 years old were collected from the Department of Radiology,affiliated Hospital of Qingdao University from 2015 to 2020,including 240males and 236females.The three-dimensional model of frontal sinus was obtained by three-dimensional reconstruction using Mimics software,and the length,width,height,left and right frontal sinus volume and total volume were measured,and the front view shape,side view shape,upper edge shape,cross-sectional shape and upper edge angle of the frontal sinus 3D model were observed and classified and coded in detail.The coding composition of frontal sinus is as follows:bilateral asymmetry index(the bilateral asymmetry index,BAI)/frontal sinus anterior view shape(left,right)/frontal sinus lateral view shape(left,right)/superior frontal sinus shape(left,right)/frontal sinus cross-sectional shape(left,right)/frontal sinus superior edge angle.Independent sample t-test was used to determine whether there was statistical significance in the difference of metrological indicators between genders,and chi-square test was used to analyze the differences in morphological indicators between sexes and sides.Calculate the distribution frequency of each morphological feature.The degree of variation of each index and the cumulative individual identification probability of 6 indexes were calculated,and the correlation among coding indexes was analyzed by Cramers V test.50samples were randomly selected for re-examination and coding,and compared with the frontal sinus coding database to verify the recognition efficiency of the coding system.Results The length,width,height,left and right frontal sinus volume and total volume of frontal sinus were significantly different between sexes(P<0.05),and the measured values of male were higher than those of female.Chi-square test showed that there was significant difference in morphological indexes between sexes(P<0.05),but there was no significant difference between left and right sides(P>0.05).The distribution frequency of each morphological feature ranges from 0.008 to 0.622,the variation degree of 6 indexes ranges from 0.56102 to 0.86014,and the cumulative individual identification probability of 6 indexes is 0.99990.There are more than 1.65888×10~8combinations of 6indexes,and the probability of forming the same code is 9.62842×10~5.A total of 467frontal sinus codes were obtained,of which 457 unique codes appeared only once,9codes appeared twice,1 code appeared 3 times,and the coding diversity was 98.1%.The correlation coefficient of each index is between 0.299 and 0.611.One of the verification samples did not match successfully.Conclusion The morphology of frontal sinus is highly specific,and the observed features have high variability among different individuals,which is a good index for individual identification.The coding method based on frontal sinus morphology has better individual recognition efficiency.It can be used as an auxiliary means of individual identification in forensic medicine.