Recognition by forensic facial approximation: case specific examples and empirical tests

The skeletal remains of one individual found near Adelaide in 1994, although not known at the time, were the first evidence of what was to be a serial killing reported to have resulted in the highest casualty list to date in Australia (12 victims). Since the usual methods of identification could not be used or were unsuccessful on these remains, facial approximations were produced and advertised over the 4-year period following their discovery, in an attempt to help to identify them. However, no identification was made. In 1999, the remains were reported to be identified by radiographic comparison. Approximately 3 months before this identification was made, another facial approximation was produced by the first author (CNS), but this face was never advertised in the media. Although rarely reported in the literature, this paper provides an example where facial approximation methods were not successful in a forensic scenario. The paper also reports on empirical tests of the facial approximation created by the first author to determine if this facial approximation might have been useful had it been advertised. The results provide further evidence that high resemblance of a facial approximation to the target individual does not indicate recognizability, as the facial approximation was poorly recognized even though it bore good resemblance to the target individual. The usefulness of facial approximation techniques is discussed within the context of this case and more broadly. Methods used to assess the accuracy of facial approximations are also discussed and further evaluated.     

Building faces from dry skulls: are they recognized above chance rates?

Methods of facial approximation have successfully aided the identification of deceased individuals. Successes may be due to either accurate facial approximation techniques or chance. This study aims to determine if any of 16 facial approximations, built using standard techniques, are sufficiently accurate to produce correct identifications of target individuals above chance. Four skulls were approximated using four commonly used methods of facial approximation. The resulting 16 facial approximations were judged by 37 assessors of varying ages. Assessors attempted to identify the target individual of each facial approximation from a face pool of ten photographed faces. Only one facial approximation resulted in true positive identification rates above chance at statistically significant levels. It is concluded that it is rare for facial approximations to be sufficiently accurate to allow identification of a target individual above chance. Since 403 incorrect identifications were made out of 592 identification scenarios, facial approximation should be considered to be a highly inaccurate and unreliable forensic technique. These results suggest that facial approximations are not very useful in excluding individuals to whom skeletal remains may not belong. Evidence from this experiment supports suggestions by others that facial approximation should be used in forensic science when all other methods of identification have failed and only to provide tentative identification.     

3D forensic facial approximation: Implementation protocol in a forensic activity

The primary objective of this paper is to report on the successful implementation of forensic facial approximation in a real case in the forensic context. A three‐dimensional (3D) facial approximation protocol of the skull was performed with free software, applying techniques in a virtual environment that have already been consolidated in the literature. The skull was scanned with the photogrammetry technique, the digital replica was imported in the Blender software (Blender Foundation, Amsterdam) and individualized model sketches of the face were traced with the MakeHuman software (MakeHuman Org) according to the anthropological profile of the victim. The face created was imported in Blender, where it was adapted, modeled, and sculpted on the 3D skull and its soft tissue markers, using an American open‐source application of the technique in the digital environment. The face created in a virtual environment was recognized and legal identification procedures were started, resulting in the more agile delivery of the disappeared body to its next of kin. It is therefore concluded that facial approximation may not be a primary method of human identification, but it can be satisfactorily applied in the forensic field as an individual recognition resource. It has great value in narrowing the search, reducing the number of alleged victims, and leading to identification tests, therefore significantly reducing the number of genetic DNA (deoxyribonucleic acid) tests—which are considered costly for the State or Federation—and consequently reducing the waiting time before delivery of the body to its family.     

STR analysis of skeletal remains exhumed from a dry “water well” used as a massive clandestine burial in Argentina

The military dictatorship that took power in Argentina from 1976 to 1983, brought enforced disappearance and death to thousands of people in the country. “Pozo de Vargas” is a 3.5 m wide and 60 m deep dry water well used as a massive clandestine burial site in Tucumán, Argentina. The site contained commingled remains as well as partially articulated body parts. Since 2011, EAAF Forensic Genetics Laboratory (LGF-EAAF) has been processing post mortem (PM) samples exhumed from the well to achieve the intra-skeletal re-association and the identification of the bodies or body parts. 981 skeletal remains, with different states of preservation, were analyzed for STR typing using commercial available kits. Different bone elements were exhumed for intra-skeletal re-associations (PM-PM genetic comparisons): teeth, femur and tibia represented more than 65% of sampling and almost all of them were successfully re-associated; spongy bones as scapula, vertebrae and clavicles yielded worse results, although almost 80% of them were also re-associated. 871 skeletal remains (89%) were successfully typed for autosomal STR markers and re-associated into 147 unique STR profiles meaning that at least 147 different individuals were thrown into the water well. After re-association, unique STR profiles obtained from the skeletal remains were compared to ante mortem (AM) family reference database in order to identify victims (AM-PM genetic comparisons), being possible to restore 113 victims to their families. Another challenge of this identification project was the presence of biologically related victims inside the well.     

The placement of the human eyeball and canthi in craniofacial identification

An accurate understanding of the spatial relationships between the deep and superficial structures of the head is essential for anthropological methods concerned with the comparison of faces to skulls (superimposition) or the prediction of faces from them (facial approximation). However, differences of opinion exist concerning: (i) the position of the eyeball in planes other than the anteroposterior plane and (ii) the canthi positions relative to the bony orbital margins. This study attempts to clarify the above relationships by dissection of a small sample of adult human cadavers (N = 4, mean age = 83 years, s = 12 years). The most notable finding was that the eyeballs were not centrally positioned within the orbits as the more recent craniofacial identification literature expounds. Rather, the eyeballs were consistently positioned closer to the orbital roof and lateral orbital wall (by 1-2 mm on average); a finding consistent with the earlier anatomical literature. While these estimation errors are small ipsilaterally, several factors make them meaningful: (i) the orbital region is heavily used for facial recognition; (ii) the width error is doubled because the eyes are bilateral structures; (iii) the eyes are sometimes used to predict/assess other soft tissue facial structures; and (iv) the net error in facial approximation rapidly accumulates with the subsequent prediction of each independent facial feature. While the small sample size of this study limits conclusive generalizations, the new data presented here nonetheless have immediate application to craniofacial identification practice because the results are evidence based. In contrast, metric data have never been published to support the use of the central positioning guideline. Clearly, this study warrants further quantification of the eyeball position in larger samples and preferably of younger individuals.     

Assessing facial approximation accuracy: how do resemblance ratings of disparate faces compare to recognition tests?

In the past, the accuracy of facial approximations has been assessed by resemblance ratings (i.e., the comparison of a facial approximation directly to a target individual) and recognition tests (e.g., the comparison of a facial approximation to a photo array of faces including foils and a target individual). Recently, several research studies have indicated that recognition tests hold major strengths in contrast to resemblance ratings. However, resemblance ratings remain popularly employed and/or are given weighting when judging facial approximations, thus indicating that no consensus has been reached. This study aims to further investigate the matter by comparing the results of resemblance ratings and recognition tests for two facial approximations which clearly differed in their morphological appearance. One facial approximation was constructed by an experienced practitioner privy to the appearance of the target individual (practitioner had direct access to an antemortem frontal photograph during face construction), while the other facial approximation was constructed by a novice under blind conditions. Both facial approximations, whilst clearly morphologically different, were given similar resemblance scores even though recognition test results produced vastly different results. One facial approximation was correctly recognized almost without exception while the other was not correctly recognized above chance rates. These results suggest that resemblance ratings are insensitive measures of the accuracy of facial approximations and lend further weight to the use of recognition tests in facial approximation assessment.     

DNA Typing for the Identification of Old Skeletal Remains from Korean War Victims*

Abstract: The identification of missing casualties of the Korean War (1950–1953) has been performed using mitochondrial DNA (mtDNA) profiles, but recent advances in DNA extraction techniques and approaches using smaller amplicons have significantly increased the possibility of obtaining DNA profiles from highly degraded skeletal remains. Therefore, 21 skeletal remains of Korean War victims and 24 samples from biological relatives of the supposed victims were selected based on circumstantial evidence and/or mtDNA‐matching results and were analyzed to confirm the alleged relationship. Cumulative likelihood ratios were obtained from autosomal short tandem repeat, Y‐chromosomal STR, and mtDNA‐genotyping results, and mainly confirmed the alleged relationship with values over 10⁵. The present analysis emphasizes the value of mini‐ and Y‐STR systems as well as an efficient DNA extraction method in DNA testing for the identification of old skeletal remains.     

Facial approximation: globe projection guideline falsified by exophthalmometry literature

The projection of the cornea from the bony orbit has been determined, in facial approximation, by centrally locating the eyeball in the orbit and positioning the cornea so that its most anterior point falls in line with a tangent dropped from the mid-superior to the mid-inferior orbital rim. However, there appears to be no scientific evidence to justify this guideline; yet, there have been numerous studies that measure globe projection in living subjects, from the lateral orbit using an exophthalmometer. The aim of this study was to determine if the traditional facial approximation guideline is consistent with the exophthalmometry literature. MRI research shows that corneal projection is underestimated using the traditional facial approximation guideline. An underestimation is also strongly supported by statistical comparisons of globe projection measures taken using more traditional instruments (e.g., Hertel's exophthalmometer) to skull morphology (p < 0.006). It is suggested that the traditional facial approximation guideline not be used in future facial approximations since it appears to under-predict anterior globe projection by 4 mm on average. It is also suggested that average exophthalmometer values be used until more accurate and precise ways of determining globe projection have been determined.     

Forensic odontologists successfully identify tsunami victims in Phuket, Thailand

More than a year has elapsed since the seaquake in South-East Asia in December 2004, and more than 92% of the non-Thai victims have been identified. About 80% of the non-Thai victims were identified by dental information. This high success rate of dental identification in Thailand was a matter of surprise for many forensic experts. Identification based on dental information is a highly efficient, reliable and rapid procedure. The conclusions drawn from the identification of tsunami victims in Thailand were recently discussed at the 17th Meeting of the Standing Committee on Disaster Victim Identification of Interpol in Lyon, and may be used to formulate new guidelines for the identification of victims.     

Investigation of Different Ways in Which the CODIS 7.0 May be Used in Mass Disaster Identification

DNA analysis is a key method for the identification of human remains in mass disasters. Reference samples from relatives may be used to identify missing persons by kinship analysis. Different methods of applying the CODIS in disaster victim identification (DVI) were investigated. Two searches were evaluated: (i) relating family relatives to a pedigree tree (FPT) and (ii) relating unidentified human remains to a pedigree tree (UPT). A joint pedigree likelihood ratio (JPLR) and rank were calculated for each search. Both searches were similar in average JPLR and rank. In exceptional cases, namely the existence of a mutation different from the CODIS model, a nonbiological father, a mistake in STR, or incorrect profile association, the UPT search returned one true rank, whereas the FPT search returned no results. This paper suggests a novel strategy to overcome these limitations and increase efficiency in conducting identification of mass disaster victims.     

Fluvial transport of human remains in the lower Mississippi River

The Mississippi River has claimed many lives over the last several decades. A better understanding of the universal dynamics of its fluvial system can help direct the production of a predictive model regarding the transportation of human remains in the river. The model may then be applied to situations where the location and the identification of water victims are necessarily part of the recovery process. Results from the preliminary phase of a longitudinal project involving the transport of human remains in the Mississippi River are presented and represent the analyses of 233 case files of river victims. A provisional model for fluvial transport of human remains in the Mississippi River is proposed and examined. This model indicates that time in the river and distance a body travels are related. Such a model may assist in pinpointing entry location for unidentified human remains found in the river or on its banks. Further, it has the potential to provide local and regional law enforcement agencies, the United States Coast Guard, and other search and rescue organizations with primary search areas when someone is missing in the river. Other results from this study indicate that a relationship exists between the side of the river where victims enter the water and the side of the river where the remains are recovered. Finally, relationships are established between the length of time before recovery of the remains and state of preservation exhibited by those remains. A secondary benefit from this study is a database of river victims that can be used by a variety of different agencies.     

The 2018 California Wildfires: Integration of Rapid DNA to Dramatically Accelerate Victim Identification

In November 2018, Butte County, California, was decimated by the Camp Fire, the deadliest wildfire in state history. Over 150,000 acres were destroyed, and at its peak, the fire consumed eighty acres per minute. The speed and intensity of the oncoming flames killed scores of people, and weeks before the fire was contained, first responders began searching through the rubble of 18,804 residences and commercial buildings. As with most mass disasters, conventional identification modalities (e.g., fingerprints, odontology, hardware) were utilized to identify victims. The intensity and duration of the fire severely degraded most of the remains, and these approaches were useful in only 22 of 84 cases. In the past, the remaining cases would have been subjected to conventional DNA analysis, which may have required months to years. Instead, Rapid DNA technology was utilized (in a rented recreational vehicle outside the Sacramento morgue) in the victim identification effort. Sixty-nine sets of remains were subjected to Rapid DNA Identification and, of these, 62 (89.9%) generated short tandem repeat profiles that were subjected to familial searching; essentially all these profiles were produced within hours of sample receipt. Samples successfully utilized for DNA identification included blood, bone, liver, muscle, soft tissue of unknown origin, and brain. In tandem with processing of 255 family reference samples, 58 victims were identified. This work represents the first use of Rapid DNA Identification in a mass casualty event, and the results support the use of Rapid DNA as an integrated tool with conventional disaster victim identification modalities.     

Identification of two fire victims by comparative nuclear DNA typing of skeletal remains and stored umbilical tissues

We describe here our collaborative efforts in identifying 2 fatalities of a fire disaster by using a variety of identification techniques. Postmortem findings in both cases were reinforced using Short Tandem Repeat (STR) DNA technology to establish with a high degree of certainty the identities of 2 child victims. STR markers used in the present study include HUMAMEL, HUMCSFIPO, HUMTHO1, HUMvWA, HUMFES/FPS, HUMF13A01, HUMFOLP23, D8S3O6, HUMFGA, and HUMTPOX. Unambiguous identification was made possible through matching DNA profiles generated from skeletal remains with those from umbilical tissues. These tissues were kept by their mothers in accordance with a Philippine tradition and were submitted for DNA analysis. Of the DNA profiles generated from exhumed bone samples of 21 child victims, comparison with the genetic profiles of children A and B obtained from umbilical tissues showed consistent DNA matches with remains 1756 and 1758, respectively.     

Utilization of complex technologies of molecular genetics for forensic expert identification of unidentified remains of victims of terrorist acts in Moscow in 1999

This paper presents the first experience gained in Russia in expert identification practice with complex utilization of technologies of molecular genetic individualization of biological objects for solving identification tasks requiring the maximum complete armory of mutually supplementing means. This is particularly important for forensic expert personality identification in cases with many deaths, when indirect identification has to be resorted to, consisting in the use of biological samples from relatives of the victims as the identifying objects. Methodological approaches and concepts of expert studies, used for identification of the remains of people who died in terroristic acts committed in Moscow in September, 1999, are discussed. The results of this study and methodological experience notably extend the potentialities of expert evaluation as regards forensic medical identification of victims of overall disasters, terroristic acts, and war conflicts.     

Forensic Facial Approximation: Study of the Nose in Brazilian Subjects

Forensic facial approximation is an auxiliary method for human identification and allows facial recognition. The midface, that includes the nose, is vital for the recognition of a familiar face. The purpose of this study was to set hard tissue parameters to estimate nasal width, to test the method to estimate nasal width of Brazilians, and to analyze the relationship between nasal profile and facial type. A total of 246 computed tomography scans (183 females and 63 males) of adults were analyzed in Horos. Bone tissue measurements and facial type classification were performed on the skull scan. Nasal profile morphology was accessed through the tool 3D surface rendering. There was a difference around 3 mm from real to predicted nose through the method to estimate nasal width in Brazilians. So, the method may be used in forensic practice. Straight nose was associated with long face type.     

Measuring the accuracy of facial approximations: a comparative study of resemblance rating and face array methods

The success of facial approximation is thought to depend, at least in part, upon the "accuracy" of the constructed face. However, methods of accuracy assessment are varied and this range in methods may be responsible for the disparate results reported in the literature. The aim of this study was to determine if the accuracy results of one facial approximation were comparable across two different assessment methods (resemblance ratings and simultaneous face array tests using unfamiliar assessors) and if resemblance ratings co-varied with recognition responses. True-positive recognition performance from the facial approximation was poor (21%) while resemblance scores using the same facial approximation were moderately high (3 out of 5 on a five-point scale). These results are not, therefore, consistent and indicate that either different variables are being evaluated by the methods, or the same variable is being examined but with different weight/calibration. Further resemblance ratings tests of the facial approximation to three foil faces from the face array revealed that resemblance scores were similar irrespective of which face was compared, and did not closely correspond with the degree of recognition performance. This was especially the case for isolated comparisons of single faces to the facial approximation. Collectively, these results indicate that resemblance ratings are: (i) insensitive measures of a facial approximation's accuracy; and (ii) inconsistent with results of unfamiliar simultaneous face-array recognition results. These data suggest that familiar and unfamiliar recognition tests should be given increased weight in contrast to current resemblance rating tests.     

Identification of exhumed remains of fire tragedy victims using conventional methods and autosomal/Y-chromosomal short tandem repeat DNA profiling

In a fire tragedy in Manila in December 1998, one of the worst tragic incidents which resulted in the reported death of 23 children, identity could not be established initially resulting in the burial of still unidentified bodies. Underscoring the importance of identifying each of the human remains, the bodies were exhumed 3 months after the tragedy. We describe here our work, which was the first national case handled by local laboratories wherein conventional and molecular-based techniques were successfully applied in forensic identification. The study reports analysis of DNA obtained from skeletal remains exposed to conditions of burning, burial, and exhumation. DNA typing methods using autosomal and Y-chromosomal short tandem repeat (Y-STR) markers reinforced postmortem examinations using conventional identification techniques. The strategy resulted in the identification of 18 out of the 21 human remains analyzed, overcoming challenges encountered due to the absence of established procedures for the recovery of mass disaster remains. There was incomplete antemortem information to match the postmortem data obtained from the remains of 3 female child victims. Two victims were readily identified due to the availability of antemortem tissues. In the absence of this biologic material, parentage testing was performed using reference blood samples collected from parents and relatives. Data on patrilineal lineage based on common Y-STR haplotypes augmented autosomal DNA typing, particularly in deficiency cases.