Techniques in facial identification: computer-aided facial reconstruction using a laser scanner and video superimposition

Cost-effective 3D scanning and printing technologies for outer ear reconstruction: current status

Current 3D scanning and printing technologies offer not only state-of-the-art developments in the field of medical imaging and bio-engineering, but also cost and time effective solutions for surgical reconstruction procedures. Besides tissue engineering, where living cells are used, bio-compatible polymers or synthetic resin can be applied. The combination of 3D handheld scanning devices or volumetric imaging, (open-source) image processing packages, and 3D printers form a complete workflow chain that is capable of effective rapid prototyping of outer ear replicas. This paper reviews current possibilities and latest use cases for 3D-scanning, data processing and printing of outer ear replicas with a focus on low-cost solutions for rehabilitation engineering.

Facial soft tissue thickness differences among different vertical facial patterns

In forensic facial approximation, facial soft tissue thickness (FSTT) measurements play a major role. These values are affected by many factors such as ethnicity, age and sex, in addition to measurement errors. We hypothesize that an additional source of error is the lack of consideration of facial type in the assessment of FSTT norms. The purpose of this study was to: 1- evaluate the presence of significant effects of vertical facial type within the FSTT measurements in adults and 2- assess the correlations between FSTT and hard and soft tissue cephalometric measurements. The sample consisted of the lateral cephalometric radiographs of 222 adult individuals (87 males; 135 females, 23.49±6.24 years of age) with normal occlusion and balanced profiles. Hard and soft tissue cephalometric measurements were taken, in addition to FSST at 10 facial landmarks. The sample was categorized into 3 vertical pattern groups based on the MP/SN angle: hypodivergent, normodivergent and hyperdivergent. Statistical analyses included MANOVA test and Pearson moment product for associations among variables. Statistically significant effect of vertical divergence on FSTT values was limited to the levels of Stomion, Labiomentale and Pogonion and FSTT measurements were associated with measurements related to the lower face (Lm and Pog) Moderate to high correlations between mandibular length and ramus length and FSTT values related to the lower face (LL, Lm and Pog mainly) emphasize further the important role of the underlying skeleton.

Digital restoration of fragmentary human skeletal remains: Testing the feasibility of virtual reality

Experts in forensic anthropology and medicine have become gradually accustomed to examining components of the human body in the virtual workspace. While the computer-assisted approach offers numerous benefits, the interactions with digital three-dimensional biological objects are often problematic, particularly if conducted with mouse, keyboard and flat-panel screen. The study focusses on feasibility of a virtual reality (VR) system for virtual restoration of fragmentary skeletal remains. The VR system was confronted with three cases of fragmentary remains. The cases were reassembled manually by twenty participants using a HTC Vive headset combined with an in-house application A.R.T. The same task was performed using a CloudCompare software in conjunction with a desktop peripheral. The two systems were compared in terms of time efficiency, the geometric properties of the resulting restorations, and convenience of use. Restoration using the VR system took approximately half the time the desktop set-up did. The VR system also yielded a lower error rate when a severely fragmented skull was reassembled. Ultimately, although the efficiency of the reassembling was shown to be strongly dependent on the operator's experience, the use of the VR system balanced out the uneven levels of proficiency in computer graphics. The current generation of virtual reality headsets has a strong potential to facilitate and improve tasks relating to the virtual restoration of fragmented skeletal remains. A VR system offers an intuitive digital working environment which is less affected by an operator's computer skills and practical understanding of the technology than the desktop systems are.

Midsagittal facial soft tissue thickness norms in an adult Mediterranean population

The use of facial soft tissue thickness (FSTT) values is essential in forensic facial approximation. Few studies have assessed the FSTT norms in Mediterranean populations and none in the Lebanese population. The purpose of this study was to (1) present midsagittal facial FSTT norms for Lebanese adults with well-balanced faces and normal occlusion; (2) evaluate the presence of sexual dimorphism within these measurements; (3) establish a potential correlation between these measurements and (4) build on the existing database of FSTT data from previously published studies. FSTT measurements at 10 midsagittal locations were obtained from the lateral cephalometric radiographs of 87 males and 135 females (mean age of 23.49±6.24years). In additional to means, Shorth and 75-Shormax values are presented. Differences between sexes were assessed using the MANOVA test and correlations between different measurements were computed. The thinnest (3.07±0.72mm) and thickest (15.61±2.38mm) craniofacial soft tissue measurements existed at Rhinion and Subnasale, respectively. There was a statistically significant and large effect of sex on the combined FSTT variables (p<0.001; eta-squared=0.393). Males displayed larger FSTT values at all landmarks except at Glabella (p=0.162). Significant correlations were observed between almost all measurements with the highest being between the upper and lower lips (r=0.763) and between Pogonion and Gnathion (r=0.784). The descriptive values of FSTT values reported in this study add to the body of research necessary for the enhancement of facial approximation methods The uniquely strong effects of sex on FSTT measurements combined and on selected single FSTT measurements are an area for further research.

A virtual reality method for digitally reconstructing traffic accidents from videos or still images

With an increase in the number of traffic accidents and enhanced attention to the rule of law, technical appraisement to reconstruct traffic accidents is attracting increasing attention. Accident videos are important aspects in identification; however, we cannot reconstruct an accident scene onsite using video for many reasons. In this paper, we introduce a computer-based virtual reality method that can digitally reconstruct a traffic accident. This method employs accident videos to perform a three-dimensional (3D) reconstruction of accident scenes. Using video screenshots, it constructs a model of humans and vehicles in 3D space to achieve the goal of dynamic restoration. The results indicate that this method has relatively high accuracy, requires little time and is easy to use. In this paper, we analyse the sources of errors for this method and summarize the application conditions.

Nose profile morphology and accuracy study of nose profile estimation method in Scottish subadult and Indonesian adult populations

This study investigated nose profile morphology and its relationship to the skull in Scottish subadult and Indonesian adult populations, with the aim of improving the accuracy of forensic craniofacial reconstruction. Samples of 86 lateral head cephalograms from Dundee Dental School (mean age, 11.8 years) and 335 lateral head cephalograms from the Universitas Padjadjaran Dental Hospital, Bandung, Indonesia (mean age 24.2 years), were measured. The method of nose profile estimation based on skull morphology previously proposed by Rynn and colleagues in 2010 (FSMP 6:20–34) was tested in this study. Following this method, three nasal aperture-related craniometrics and six nose profile dimensions were measured from the cephalograms. To assess the accuracy of the method, six nose profile dimensions were estimated from the three craniometric parameters using the published method and then compared to the actual nose profile dimensions.

In the Scottish subadult population, no sexual dimorphism was evident in the measured dimensions. In contrast, sexual dimorphism of the Indonesian adult population was evident in all craniometric and nose profile dimensions; notably, males exhibited statistically significant larger values than females. The published method by Rynn and colleagues (FSMP 6:20–34, 2010) performed better in the Scottish subadult population (mean difference of maximum, 2.35 mm) compared to the Indonesian adult population (mean difference of maximum, 5.42 mm in males and 4.89 mm in females).

In addition, regression formulae were derived to estimate nose profile dimensions based on the craniometric measurements for the Indonesian adult population. The published method is not sufficiently accurate for use on the Indonesian population, so the derived method should be used. The accuracy of the published method by Rynn and colleagues (FSMP 6:20–34, 2010) was sufficiently reliable to be applied in Scottish subadult population.

On the integral use of foundational concepts in verifying validity during skull-photo superimposition

Often cited reliability test on video superimposition method integrated scaling face-images in relation to skull-images, tragus-auditory meatus relationship in addition to exocanthion-Whitnall's tubercle relationship when orientating the skull-image and wipe mode imaging in addition to mix mode imaging when obtaining skull-face image overlay and evaluating the goodness of match. However, a report that found higher false positive matches in computer assisted superimposition method transited from the above foundational concepts and relied on images of unspecified sizes that are lesser than 'life-size', frontal plane landmarks in the skull- and face- images alone for orientating the skull-image and mix images alone for evaluating the goodness of match. Recently, arguing the use of 'life-size' images as 'archaic', the authors who tested the reliability in the computer assisted superimposition method have denied any method transition. This article describes that the use of images of unspecified sizes at lesser than 'life-size' eliminates the only possibility to quantify parameters during superimposition which alone enables dynamic skull orientation when overlaying a skull-image with a face-image in an anatomically acceptable orientation. The dynamic skull orientation process mandatorily requires aligning the tragus in the 2D face-image with the auditory meatus in the 3D skull-image for anatomically orientating the skull-image in relation to the posture in the face-image, a step not mentioned by the authors describing the computer assisted superimposition method. Furthermore, mere reliance on mix type images during image overlay eliminates the possibility to assess the relationship between the leading edges of the skull- and face-image outlines as also specific area match among the corresponding craniofacial organs during superimposition. Indicating the possibility of increased false positive matches as a consequence of the above method transitions, the need for testing the reliability in the superimposition method adopting concepts that are considered safe is stressed.

Craniofacial Superimposition: Historical Review and Current Issues

Methodology of craniofacial superimposition has evolved dramatically since its inception in the early 20th century. Contemporary approaches involve advanced procedures of digitization, use of video cameras, warping algorithms, skull positioning devices and data derived from computed tomography, radiography, and ultrasound technology. Applications have proven most useful in exclusion and casework frequency has declined in some regions in recent years as molecular approaches to identification have become more widely available. Additional research is needed to clarify the probabilities involved and to facilitate comparisons when the technique is employed.

Study on the performance of different craniofacial superimposition approaches (I)

As part of the scientific tasks coordinated throughout The 'New Methodologies and Protocols of Forensic Identification by Craniofacial Superimposition (MEPROCS)' project, the current study aims to analyse the performance of a diverse set of CFS methodologies and the corresponding technical approaches when dealing with a common dataset of real-world cases. Thus, a multiple-lab study on craniofacial superimposition has been carried out for the first time. In particular, 26 participants from 17 different institutions in 13 countries were asked to deal with 14 identification scenarios, some of them involving the comparison of multiple candidates and unknown skulls. In total, 60 craniofacial superimposition problems divided in two set of females and males. Each participant follow her/his own methodology and employed her/his particular technological means. For each single case they were asked to report the final identification decision (either positive or negative) along with the rationale supporting the decision and at least one image illustrating the overlay/superimposition outcome. This study is expected to provide important insights to better understand the most convenient characteristics of every method included in this study.

A facial reconstruction and identification technique for seriously devastating head wounds

Many authors have focused on facial identification techniques, and facial reconstructions for cases when skulls have been found are especially well known. However, a standardized facial identification technique for an unknown body with seriously devastating head injuries has not yet been developed. A reconstruction and identification technique was used in 7 cases of accidents involving trains striking pedestrians. This identification technique is based on the removal of skull bone fragments, subsequent fixation of soft tissue onto a universal commercial polystyrene head model, precise suture of dermatomuscular flaps, and definitive adjustment using cosmetic treatments. After reconstruction, identifying marks such as scars, eyebrows, facial lines, facial hair and partly hairstyle become evident. It is then possible to present a modified picture of the reconstructed face to relatives. After comparing the results with photos of the person before death, this technique has proven to be very useful for identifying unknown bodies when other identification techniques are not available. This technique is useful for its being rather quick and especially for its results.

Is characterizing the digital forensic facial reconstruction with hair necessary? A familiar assessors' analysis

BACKGROUND

In the international scientific literature, there are few studies that emphasize the presence or absence of hair in forensic facial reconstructions. There are neither Brazilian studies concerning digital facial reconstructions without hair, nor research comparing recognition tests between digital facial reconstructions with hair and without hair. The miscegenation of Brazilian people is considerable. Brazilian people, and, in particular, Brazilian women, even if considered as Caucasoid, may present the hair in very different ways: curly, wavy or straight, blonde, red, brown or black, long or short, etc. For this reason, it is difficult to find a correct type of hair for facial reconstruction (unless, in real cases, some hair is recovered with the skeletal remains).

AIMS AND METHODS

This study focuses on the performance of three different digital forensic facial reconstructions, without hair, of a Brazilian female subject (based on one international database and two Brazilian databases for soft facial-tissue thickness) and evaluates the digital forensic facial reconstructions comparing them to photographs of the target individual and nine other subjects, employing the recognition method. A total of 22 assessors participated in the recognition process; all of them were familiar with the 10 individuals who composed the face pool.

RESULTS AND CONCLUSIONS

The target subject was correctly recognized by 41% of the 22 examiners in the International Pattern, by 32% in the Brazilian Magnetic Resonance Pattern and by 32% in the Brazilian Fresh Cadavers Pattern. The facial reconstructions without hair were correctly recognized using the three databases of facial soft-tissue thickness. The observed results were higher than the results obtained using facial reconstructions with hair, from the same skull, which can indicate that it is better to not use hair, at least when there is no information concerning its characteristics.

An investigation into the accuracy and reliability of skull-photo superimposition in a South African sample

One of the aims of forensic science is to determine the identities of victims of crime. In some cases the investigators may have ideas as to the identities of the victims and in these situations, ante mortem photographs of the victims could be used in order to try and establish identity through skull-photo superimposition. The aim of this study was to evaluate the accuracy of a newly developed digital photographic superimposition technique on a South African sample of cadaver photographs and skulls. Forty facial photographs were selected and for each photo, 10 skulls (including the skull corresponding to the photo) were used for superimposition. The investigator did not know which of the 10 skulls corresponded to the photograph in question. The skulls were scanned 3-dimensionally, using a Cyberware™ Model 3030 Colour-3D Scanhead scanner. The photos were also scanned. Superimposition was done in 3D Studio Max and involved a morphological superimposition, whereby a skull is superimposed over the photo and assessed for a morphological match. Superimposition using selected anatomical landmarks was also performed to assess the match. A total of 400 skull-photo superimpositions were carried out using the morphological assessment and another 400 using the anatomical landmarks. In 85% of cases the correct skull was included in the possible matches for a particular photo using morphological assessment. However, in all of these cases, between zero and three other skulls out of 10 possibilities could also match a specific photo. In the landmark based assessment, the correct skull was included in 80% of cases. Once again, however, between one and seven other skulls out of 10 possibilities also matched the photo. This indicates that skull-photo superimposition has limited use in the identification of human skeletal remains, but may be useful as an initial screening tool. Corroborative techniques should also be used in the identification process.

A local technique based on vectorized surfaces for craniofacial reconstruction

In this paper, we focus on the automation of facial reconstruction. Since they consider the whole head as the object of interest, usual reconstruction techniques are global and involve a large number of parameters to be estimated. We present a local technique which aims at reaching a good trade-off between bias and variance following the paradigm of non-parametric statistics. The estimation is localized on patches delimited by surface geodesics between anatomical points of the skull. The technique relies on a continuous representation of the individual surfaces embedded in the vectorial space of extended normal vector fields. This allows to compute deformations and averages of surfaces. It consists in estimating the soft-tissue surface over patches. Using a homogeneous database described in [31], we obtain results on the chin and nasal regions with an average error below 1mm, outperforming the global reconstruction techniques.

Construction and analysis of a head CT-scan database for craniofacial reconstruction

This paper is devoted to the construction of a complete database which is intended to improve the implementation and the evaluation of automated facial reconstruction. This growing database is currently composed of 85 head CT-scans of healthy European subjects aged 20-65 years old. It also includes the triangulated surfaces of the face and the skull of each subject. These surfaces are extracted from CT-scans using an original combination of image-processing techniques which are presented in the paper. Besides, a set of 39 referenced anatomical skull landmarks were located manually on each scan. Using the geometrical information provided by triangulated surfaces, we compute facial soft-tissue depths at each known landmark positions. We report the average thickness values at each landmark and compare our measures to those of the traditional charts of [J. Rhine, C.E. Moore, Facial Tissue Thickness of American Caucasoïds, Maxwell Museum of Anthropology, Albuquerque, New Mexico, 1982] and of several recent in vivo studies [M.H. Manhein, G.A. Listi, R.E. Barsley, et al., In vivo facial tissue depth measurements for children and adults, Journal of Forensic Sciences 45 (1) (2000) 48-60; S. De Greef, P. Claes, D. Vandermeulen, et al., Large-scale in vivo Caucasian facial soft tissue thickness database for craniofacial reconstruction, Forensic Science International 159S (2006) S126-S146; R. Helmer, Schödelidentifizierung durch elektronische bildmischung, Kriminalistik Verlag GmbH, Heidelberg, 1984].

Assessment of the accuracy of three-dimensional manual craniofacial reconstruction: a series of 25 controlled cases

The aim of this work was to estimate the accuracy of craniofacial reconstruction (CFR), from a series of 25 controlled cases. Three protocols of blind CFRs (exhibiting an increasing complexity from A to C) were assessed in this paper, allowing comparison of the CFR with the actual face of the deceased. The whole results showed that an excellent, or good, to middle resemblance (between the blind CFR and the actual face of the subject) was reached in 9 out of 25 cases, but the success gradually increased from the A to the C protocol of CFR, reaching six cases out of eight in the latter. Statistical comparison of measurements (between the blind CFR and the actual face) was also achieved, revealing that some anthropological distances were constantly underestimated or overestimated. This experiment shows that a thorough anthropological, odontological, and X-ray analysis is indispensable before performing a CFR, and these encouraging results justify further efforts of research in this field.

Use of hand-held laser scanning in the assessment of craniometry

The intra- and inter-examiner reliability was evaluated for hand-held 3D laser scanning technology when it was combined with localization of landmarks for craniometry. The data from the laser surface scanning were compared with those of conventional direct measuring. Using thirty unidentified skulls requested for individual identification, measurements were taken of the line distances from lambda to 26 landmarks, and also for seven breadth parameters. For the laser surface scanning, two examiners performed replicate measurements with an interval of 1 week. In the conventional direct measuring, the first examiner took replicate measurements with a 1-week interval. To assess intra- and inter-examiner reliabilities, the intraclass correlation coefficient was used. Analysis of variance with repeated measures for each parameter was performed to compare the conventional method with the 3D scanning method. Both the 3D scanning and conventional methods showed excellent intra-examiner reliabilities, and the 3D laser scanning method also showed excellent inter-examiner reliability. A statistical difference between the two examiners was found only in nasal breadth in the 3D laser scanning method. There was no significant difference between the two measuring methods, though the 3D laser scanning method tended to give a slightly lower reading. Collectively, the 3D laser scanning method with point localization is a useful method with excellent reliability, and it can replace the conventional direct measuring method in craniometry.

Computerized craniofacial reconstruction using CT-derived implicit surface representations

In forensic craniofacial reconstruction, facial features of an unknown individual are estimated from an unidentified skull, based on a mixture of experimentally obtained guidelines on the relationship between soft tissues and the underlying skeleton. In this paper, we investigate the possibility of using full 3D cross-sectional CT images for establishing a reference database of densely sampled distances between the external surfaces of the skull and head for automated craniofacial reconstruction. For each CT image in the reference database, the hard tissue (skull) and soft tissue (head) volumes are automatically segmented and transformed into signed distance transform (sDT) images, representing for each voxel in this image the Euclidean distance to the closest point on the skull and head surface, respectively, distances being positive (negative) for voxels inside (outside) the skull/head. Multiple craniofacial reconstructions are obtained by first warping each reference skull sDT maps to the target skull sDT using a B-spline based free form deformation algorithm and subsequently applying these warps to the reference head sDT maps. A single reconstruction of the target head surface is defined as the zero level set of the arithmetic average of all warped reference head sDT maps, but other reconstructions are possible, biasing the result to subject specific attributes (age, BMI, gender). Both qualitative and quantitative tests (measuring the similarity between the 3D reconstructed and corresponding original head surface) on a small (N = 20) database are presented to proof the validity of the concept.

Craniofacial reconstruction using a combined statistical model of face shape and soft tissue depths: methodology and validation

Forensic facial reconstruction aims at estimating the facial outlook associated with an unidentified skull specimen. Estimation is generally based on tabulated average values of soft tissue thicknesses measured at a sparse set of landmarks on the skull. Traditional 'plastic' methods apply modeling clay or plasticine on a cast of the skull, approximating the estimated tissue depths at the landmarks and interpolating in between. Current computerized techniques mimic this landmark interpolation procedure using a single static facial surface template. However, the resulting reconstruction is biased by the specific choice of the template and no face-specific regularization is used during the interpolation process. We reduce the template bias by using a flexible statistical model of a dense set of facial surface points, combined with an associated sparse set of skull-based landmarks. This statistical model is constructed from a facial database of (N = 118) individuals and limits the reconstructions to statistically plausible outlooks. The actual reconstruction is obtained by fitting the skull-based landmarks of the template model to the corresponding landmarks indicated on a digital copy of the skull to be reconstructed. The fitting process changes the face-specific statistical model parameters in a regularized way and interpolates the remaining landmark fit error using a minimal bending thin-plate spline (TPS)-based deformation. Furthermore, estimated properties of the skull specimen (BMI, age and gender, e.g.) can be incorporated as conditions on the reconstruction by removing property-related shape variation from the statistical model description before the fitting process. The proposed statistical method is validated, both in terms of accuracy and identification success rate, based on leave-one-out cross-validation tests applied on the facial database. Accuracy results are obtained by statistically analyzing the local 3D facial surface differences of the reconstructions and their corresponding ground truth. Identification success rate is obtained by comparing, based on correlation, Euclidean distance matrix (EDM) signatures of the reconstructed and the original 3D facial surfaces in the database. A subjective identification success rate is quantified based on face-pool tests. Finally a qualitative comparison is made between facial reconstructions of a real-case skull, based on two typical static face models and our statistical model, showing the shortcomings of current face models and the improved performance of the statistical model.

Skeletal remains of Dr. Eugenio Antonio Berríos Sagredo

It is often noted that even a well-designed osteological technique may not provide accurate results when applied to single forensic cases. Case studies are ideal to test if this concern is valid, and forensic anthropology is a testing ground for applying a population based standard to individual skeletal remains. Secondly, the increasing role anthropologists have played in forensic sciences has aided the medicolegal disciplines in a number of ways. For example, identification of skeletal remains is now more accurate than ever before. Many of these cases have brought perpetrators to court for justice. The purpose of this paper is to use osteological techniques to analyze skeletal remains and make a positive identification. The victim was found partially buried in the sand near El Pinar, Uruguay in 1995. The analysis indicated that the victim was a 45-year old, white, male who was about 170cm tall. Based on preliminary evidence that the victim might be Dr. Eugenio Antonio Berríos Sagredo, a digital superimposition was made using the victim's photograph and the unknown skull. This examination revealed that the skull corresponded consistently with the individual in the photograph. Results were supported by the fact that personal belongings, such as a medal and wrist watch, also pointed to the same individual. Dental records and radiographs when made available later also indicated the same identity. Dr. Berríos was accused of making nerve gas during the dictatorial regime of former Chilean President General Augusto Pinochet. It was also alleged that he made bombs that killed a Spanish diplomat in his laboratory and a Chilean diplomat in Washington, DC. Many complex techniques are often needed to make a positive identification and such was the case for this study. Because of the nature of anthropology as a holistic discipline, such complexity is an integral part of human biology and behavior and can be used successfully in the forensic sciences and medicolegal investigations.

Facial reconstruction using 3-D computer graphics

Facial reconstruction using 3-D computer graphics is being used in our institute as a routine procedure in forensic cases as well as for skulls of historical and archaeological interest. Skull and facial data from living subjects is acquired using an optical laser scanning system. For the production of the reconstructed image, we employ facial reconstruction software which is constructed using the TCL/Tk scripting language, the latter making use of the C3D system. The computer image may then be exported to enable the production of a solid model, employing, for example, stereolithography. The image can also be modified within an identikit system which allows the addition of facial features as appropriate.

The application of volume deformation to three-dimensional facial reconstruction: a comparison with previous techniques

Facial reconstruction has been widely criticised for its subjective nature and is thus often viewed as a last resort. Recent computer-aided reconstructions, which claim to be objective, may be subject to similar errors. To date, both manual reconstructions and computer-aided techniques have been limited to using the same tissue depth data reference tables. We propose that it is not solely the accuracy of the soft tissue depth data, but the sparsity of landmarks, which contributes to a lack of understanding of how soft tissue changes between the landmarks. The authors feel that a new approach to facial reconstruction using a computer graphics technique known as volume deformation addresses some of the problems encountered in the past. A review of previous methods is provided with discussion of the strengths and limitations of these techniques. In addition, areas are highlighted where the new deformation-based approach may offer possible solutions.