Rajshekar_whole_thesis.pdf (1.85 MB)
Application of 3D scanning technology in forensic investigation of bite-marks
thesisposted on 2023-05-27, 10:43 authored by Rajshekar, M
Forensic investigation of bite-marks on humans has the potential to provide evidence that can be used to identify the perpetrator of a bite. Bite-mark evidence has been used in legal proceedings since 1692 but, recently, bite-mark analysis has been subject to substantial criticism. In the USA, there have now been 24 cases involving human perpetrators in which convictions based on bite-mark analysis have been overturned as a result of DNA evidence, including cases in which the defence experts had testified that the suspect dentitions did not match the bite-marks. The overturn of these convictions and the criticisms that followed have led to calls to halt the use of bite-marks as evidence until its scientific credibility can be established, and even to recommendations to discontinue altogether the use of bite-mark evidence in criminal investigations in the US. To address the fundamental limitations of bite-mark analysis, the aims of this thesis were to estimate the frequency of occurrence of bites, propose the use of 3D imaging technology as an approach to overcome limitations of current methods of bite-mark analysis, investigate the reliability and validity of measurements of landmark dental features made using 3D imaging, and examine the accuracy of matching 3D images of bite-marks to 3D images of candidate dentitions. This thesis is made up of four key studies. They are summarized below: Study 1 An important first step was to undertake an assessment of the public health implications of bites inflicted on humans by estimating the frequency of occurrence of the injuries and deaths caused. For information on bites perpetrated by humans on humans, searches and data requests were made on public access information websites and of the Australian Institute of Health and Welfare (AIHW) respectively. The AIHW was unable to provide age and sex specific information on bites caused by humans on other humans. The author's investigations revealed that there is incomplete and fragmented information on bites perpetrated by humans on humans most commonly in cases of sexual assault and child abuse. Instead the author turned their attention to bites perpetrated by other biting animals, and specifically by dogs because dog-bites are the most common. To determine the extent of the problem, the incidence of public sector hospitalizations resulting from dog-bite related injuries in Australia during the period 2001-2013 was estimated. The principal finding was that on average, 2061 persons were hospitalized each year during that period for treatment for dog-bite injuries at an annual rate of 12.39 (95% CI 12.25, 12.53) per 100,000. The highest annual rates of 25.95 (95% CI 25.16, 26.72) and 18.42 (95% CI 17.75, 19.07) per 100,000 were for age groups 0-4 years and 5-9 years respectively. Rates of recorded events increased over the study period and reached 16.15 (95% CI 15.78, 16.52) per 100,000 during 2011-13. This study was the first national study to report the incidence of hospitalization for injuries due to dog-bites for an extended period with complete coverage of all public hospitals in Australian states and territories. The findings add considerably to what is known about the public health problem of dog-bite injuries in Australia. Study 2 Recently, the scientific basis of bite-mark analysis has been questioned. The most robust of those criticisms came from the National Academy of Science (NAS). In its report published in 2009, the NAS identified three fundamental limitations of bite-mark analysis. These were that the uniqueness of human dentitions was yet to be determined, that the ability of the human skin to retain faithfully the impression of the biting dentition was yet to be ascertained, and that comprehensive steps to minimise and quantify all other sources of error in matching a bite-mark to a suspect dentition were yet to be undertaken. To address these fundamental limitations, the author recommends the use of 3D imaging techniques in bite-mark analysis. The new generation of portable, non-invasive, hand-held intra-oral 3D scanners, that are currently used as an alternative to conventional dental impression materials in clinical dentistry, have made the process of acquiring dental impressions faster and easier. The 3D scanning permits the imaging of bite-marks as well as the imaging of biting dentitions of suspected perpetrators in 3 dimensions with high resolution, and would allow researchers to compile large databases of virtual images of dentitions of biting animals for quantifying population variation. In addition, 3D scanning would make it feasible to compare a large number of landmark features when matching the scan of a bite-mark to the scan of a candidate dentition. This technology makes it possible to investigate the sources of error and quantify that error, and thereby has the potential to remove or at least reduce error including by limiting subjectivity associated with human judgement. Study 3 To investigate the reliability and validity of measurements made with an intra-oral 3D scanner, two raters each measured 84 tooth and 26 arch features of 50 sets of upper and lower human dental casts first using digital hand-held callipers and second using the measuring tool provided with the ZFX IntraScan intra-oral 3D scanner applied to 3D images of the dental casts. The measurements were repeated at least one week later. Reliability and validity were quantified concurrently by calculation of intra-class correlation coefficients (ICC) and standard errors of measurement (SEM). The measurements of 110 landmark features of human dental casts made using the intra-oral 3D scanner were virtually indistinguishable from measurements of the same features made using conventional hand-held callipers. The difference of means as a percentage of the average of the measurements by each method ranged between 0.030% and 1.134%. The inter-method SEMs ranged between 0.037% and 0.535%, and the inter-method ICCs ranged between 0.904 and 0.999, for both the upper and the lower arches. The inter-rater SEMs were one-half, and the intra-method/rater SEMs were one-third, of the inter-method values. This study demonstrated that the ZFX Intrascan intra-oral 3D scanner with its virtual on-screen measuring tool is a reliable and valid method for measuring the key features of human dental casts. Study 4 The aim of study 4 was to assess the accuracy of matching 3D images of 3D impressions of dental arches with 3D images of candidate dentitions. In this proof-of-concept study, dog dental arches were used as a model because demonstration of success in matching to dentitions with substantial variation due to inter-breed differences between dogs was a logical starting point. A further consideration was that there was greater access to dog dentitions than to human dentitions. 3D images of dog dental arches and their impressions were recorded using the intra-oral 3D scanner A single rater measured and re-measured 79 landmark dental features on each of the 3D images of 40 upper and the lower dog dental arches (positive images) using the virtual onscreen measuring tool provided with the intra-oral 3D scanner. This was repeated for 3D images of the impressions of dog dental arches in modelling clay (negative images). Measurements extracted from the images were used in an attempt to match the negative images to positive images. The measurements of 79 landmark features of the dog dental arches were nearly identical to measurements of the same features made on the impressions of dog dental arches on clay. The intra-rater intra-class correlation coefficients (ICCs) were greater than 0.937, and the intra-rater SEMs ranged between 0.041 and 0.076 for both the upper and the lower arches. Using the measurements of all 79 features, or even just those of anterior dentition, there was a 100% accuracy in matching the negative images to positive images of the dentitions. Nevertheless, some features contributed to the overall match rate more than the others. This study demonstrates that the portable intra-oral 3D scanner can record impressions of dental features with sufficient accuracy to allow identification of the dog-dentition that caused an undistorted bite-mark. In conclusion, bite-mark evidence has the potential to provide supporting evidence to build a case against the human or animal perpetrator of a bite. Even when DNA evidence is available, and that is always not the case, supporting evidence from finger-prints and bite-marks helps to build a more compelling case. However, for bite-mark evidence to have probative value, it is necessary to address scientific criticisms of the methods used in obtaining it. In this thesis,this author proposes the use of intra-oral 3D scanning principally because of its potential for limiting and quantifying error. The reliability and validity of measurements of human dentitions has been established and a proof-of-concept of the accuracy of matching bite-marks with dentitions has been provided. A framework for future research is proposed, with recommendations for addressing the other limitations of bite-mark analysis. Specifically, the author proposes the establishment of databases of scanned images of dentitions of biting animals and map a path forward in the investigation of distortion in bite-marks.
Rights statementCopyright 2017 the author Chapter 2 appears to be the equivalent of a post-print version of an article published as: Rajshekar, M., Blizzard, L., Julian, R., Williams, A., Tennant, M., Forrest, A., Walsh, L. J., Wilson, G., 2017. The incidence of public sector hospitalisations due to dog bites in Australia 2001‚Äö-2013, Australian and New Zealand journal of public health, 41(4), 377-380. The article is published as an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. https://creativecommons.org/licenses/by-nc-nd/4.0/ Chapter 4 appears to be the equivalent of a post-print version of an article published as: Rajshekar, M., Julian, R., Williams, A. M., Tennant, M., Forrest, A., Walsh, L. J., Wilson, G., Blizzard, L., 2017. The reliability and validity of measurements of human dental casts made by an intra-oral 3D scanner, with conventional hand-held digital callipers as the comparison measure, Forensic science international, 278, 198-204