FIGURE 4 - uploaded by Carl N Stephan
Content may be subject to copyright.
Determination of the nose projection. a, The maxillary regions used to estimate the direction of the lower nose tangent outlined on either side of the anterior nasal spine; b, Profile view of the nasal profile with the two tangents represented in the wax; c, The outline of the infero-lateral rim of the nasal aperture, which was used to approximate the outline of the tip of the nose; d, A profile view of the constructed nose indicating: i) the distal nasal bone segment used to determine the upper tangent direction indicated-note that only the end of the nasal bone distal to the black bar was used in the estimation; and ii) the lateral nasal aperture outline (black line) used to approximate the tip of the nose-this has been rotated from c to create the nose tip.
Contexts in source publication
Context 1
... profile outline of the nose at the bridge was strictly determined by the nasal bones, the soft tissue here conforming closely to the measurement at rhinion. If the upper border of the nasal aperture was bent slightly inwards then the nasal profile, in the area of the nasal cartilages, had a wavy appearance (Figures 3-4). The profile outline of the dorsum and tip of the nose roughly corresponded to the lateral curve of the nasal aperture as seen in a frontal view (also see Balueva in Conant 2003; Figure 4). If the infero-lateral rim of the nasal aperture was rounded, then this too was mirrored in the tip of the nose. With regards to the lips, stomion was represented half way down the central incisors, and the vermillion border of the upper lip was set at the same height as the lower edge of the alveolar rim ( Figure 3). All these features were determined relative to the Frankfurt ...
Context 2
... profile line over the nasal aperture was supported by first "blocking-in" the aperture, initially with paper and then by a covering layer of mastic (as for the temporalis muscle). The "blocking-in" should be deep enough so that the anterior portions of the crista conchalis remain exposed on each side. The nose and mouth form were modeled following the two- dimensional graphic template, which was generated prior to undertaking the 3D reconstruction. More specifically, the projection of the nose was determined by placing two tangents: one following the general direction, but the very distal end (approximately last 2 mm, but exact distance was never specified) of the nasal bones. Note here that the lateral parts of the nasal bones adjacent to the median profile may be used instead of the morphology expressed directly along the median plane). The other tangent followed the general direction of the left or right floor of the anterior part of the nasal aperture (maxillary bone) laterally adjacent to the anterior nasal spine and vomer bone (Figure 4). In contradiction to the popular English literature (Rynn, Wilkinson 2006 and The Face Finder, the lower tangent did not necessarily follow the projection of the anterior nasal spine (Ullrich, Stephan 2011), but rather b, Nine of the landmarks along the profile where average soft tissue depths were used by Gerasimov (black dots). Note here that the soft tissue depth at glabella has been reduced to accompany the reduced bony relief at this region in this individual; c, Representation of the profile line on the skull using the wax mastic, constructed by first using the 2D graphic profile reconstruction at (a) as a template. the floor of the nasal aperture. Both HU and CS find this guideline difficult to implement in practice, since the floor of the nasal cavity is often undulating, especially at its anterior ...
Citations
... In facial approximation, similar applies, however, instead of using an antemortem reference photograph, mean FSTTs are used as a guide to how much soft tissue should be added to the skull to approximate an individual's face [2,3]. This applies no matter which facial approximation method is used, including so-called "Russian," "American," and "Combination" methods, as all methods, including Gerasimov's techniques, use mean FSTTs [3][4][5][6][7]. ...
This year (2023) marks 140 years since the first publication of a facial soft tissue thickness (FSTT) study. Since 1883, a total of 139 studies have been published, collectively tallying > 220,000 tissue thickness measurements of > 19,500 adults. In just the last 5-years, 33 FSTT studies have been conducted. Herein, we add these data (plus an additional 20 studies) to the 2018 T-Table to provide an update of > 81,000 new datapoints to the global tallied facial soft tissue depths table. In contrast to the original 2008 T-Table, some notable changes are as follows: increased FSTTs by 3 mm at infra second molar (ecm2–iM2ʹ), 2.5 mm at gonion (go–goʹ), 2 mm at mid-ramus (mr–mrʹ), and 1.5 mm at zygion (zy–zyʹ). Rolling grand means indicate that stable values have been attained for all nine median FSTT landmarks, while six out of nine bilateral landmarks continue to show ongoing fluctuations, indicating further data collection at these landmarks holds value. When used as point estimators for individuals with known values across 24 landmarks (i.e., C-Table data), the updated grand means produce slightly less estimation error than the 2018 T-Table means (3.5 mm versus 3.6 mm, respectively). Future efforts to produce less noisy datasets (i.e., reduce measurement and sampling errors as much as possible between studies) would be useful.
... His method included sculpture of the main mastication muscles (masseter and temporalis), while the rest of the face was constructed based on average soft tissue thickness measurements. His technique was later became known as the Russian method (Ullrich & Stephan, 2016). Wilton Marion Krogman (1903-1987, Betty Pat Gatliff, and Clyde Snow are the founders of the American method. ...
The human face is a highly specialized anatomical entity. Its appearance is a result of evolutionary processes that include biomechanical, physiological, and social influences. The facial approximation is based on the relevance of skeletal and soft tissue anatomy and aims to reconstruct a person's face from the anatomical features of the skull. In forensics, the technique focuses on the estimation of the facial characteristics of unidentified individuals. In archaeological contexts, it has been widely used for the representation of hominins, prehistoric and historic periods individuals. For the facial approximation of the Petralona man, we digitized a cast of the original skull with an Artec Spider 3D scanner of metrological accuracy up to 0.05mm. A digital model of the Mauer mandible (Germany), the holotype of Homo heidelbergensis, was used to virtually complement the skeletal anatomy of the face. Cinema4D (Maxon) was used to place soft tissue depth markers according to standard nomenclature. The digital reconstruction of facial anatomy and skin texture was developed in ZBrush (Pixologic). The facial features were approximated with the use of techniques that are based on anatomical and/or statistical evidence. The facial approximation of the Petralona skull aims to disseminate scientific information and contribute to the public understanding of evolutionary science.
... As the discipline of FFA shifts towards an automated and computerized 3D approach, large databanks of population-specific data must be created to create statistical models for predicting soft tissue facial features based on hard tissue structures. Ear shape has been commonly ignored in the literature and standard casts are used instead [2,43], but for an automatic facial approximation, an accurate estimate of the ear based on population affinity, sex, and age-specific guidelines is required. ...
Research on how to reliably reconstruct the shape of the ear for facial approximations is limited, especially in countries such as South Africa where standard ear casts are still used in manual methods. To improve objectivity, computer aided methods are being developed for facial approximations – which require extensive population specific datasets for facial feature morphology.
This study aims to assess variations in the shape of the ear and the underlying external auditory meatus (EAM) through the analysis of cone-beam computed tomography (CBCT)
scans of 40 black South Africans (males n = 17; females n = 23) and 76 white South Africans (males n = 29; females n = 47) between the ages of 18 and 90 years. Shape data was collected by placing 19 capulometric landmarks on the 3D reconstructions of the ear and 46 sliding craniometric landmarks along the EAM.
Geometric morphometric analysis revealed highly significant variation in ear shape between groups for population affinity
(p-value = 0.001), while sex and age were only significant between the white South Africans (p-value < 0.05).
Only population affinity significantly influenced shape in the EAM (p-value = 0.001), and both the ear and EAM showed significant levels of symmetry (p-value = 0.007). While an ear will never be exactly recreated, basing facial estimates on the decedent’s biological profile can lead towards the highest possible accuracies. For the ear shape specifically, sex and age will not be a priority when creating predictive models, but population affinity will greatly influence the output.
... However, reconstruction criteria for eyes, nose, and mouth, which play an important role in personal identification and facial recognition [4,22,23], have been excluded from the conventional measuring marker points of 3D facial reconstruction due to their complex morphometry, large individual differences, and difficulties in obtaining scientific criteria for measurement. Furthermore, these features are also subject to change depending on the degree of skin tension in facial expressions. ...
... In the artistic phase, while the setting position of the eyeballs has been analyzed [24][25][26], methods to reconstruct the eyelids, external nose, and lip dimensions remain controversial [22,23,27]. Considering these technical difficulties, it is better to obtain scientific measurement data for 3D facial reconstruction using the external nose because of low-mobility cartilage framework and thin skin covering [28,29]. ...
The eyelids, external nose, and lips play an important role in individual identification and facial recognition; however, they are excluded from tissue marker points, and are reconstructed based on generic methods for 3D facial reconstruction or facial approximation. Therefore, this study focused on nasal dimensions and evaluated whether Krogman’s widely used formula for estimating the dimensions of an external nose, regardless of sex, race, and body physique, can be applied to Japanese adults. A total of 146 postmortem CT images of Japanese adult cadavers (64 males and 82 females, aged 58–105 years old) were retrospectively analyzed. The total nasal projection (TNP) among Japanese adults was estimated using the formula, TNP = 1.9 × the anterior nasal spine projection (ANSP) + the mid-philtrum depth (MPD), which differed significantly from the coefficient (3.0) in the conventional formula, regardless of sex, race, and body physique, and therefore needed modification for Asians. Although there was no positive relationship between the total nasal width (TNW) and the maximum width of the anterior nasal aperture (ANAW), the TNW could be estimated by adding soft tissue that varies by sex and body physique to both sides of the nearly constant ANAW. Therefore, we determined a simple and practical formula to estimate nasal dimensions among Japanese adults for conventional 3D facial reconstruction and manual 3D facial sculpture.
... Most computer-based reconstruction method uses the American [15] method approach. Another common method is Gerasimov [16]. Both rely on landmark positioning and simulate the presence of soft tissue. ...
Computer‐aided craniofacial reconstruction (CFR) is a process that aims to estimate facial impressions based on skull remains. It mimics the conventional method with a conceptual model‐based framework. The existing problems in CFR are that landmark annotation is expert‐dependent, landmark processing in the 3D domain has volumetric challenges, and a method based on a population's morphological characteristics (templates). A framework with three stages is proposed: Building a craniofacial model, automatic landmark detection, and surface deformation. Machine learning is deployed to draw local surface correlations as landmarks and automatically detects their position. The local surface context is extracted using the Surface Curvature Feature (SCF) as a 3D descriptor. Using a cluster‐based filter, the average distance (to the ground truth) of the top 20 points is 0.0326 units. Cluster‐based filters are better than mass‐radius‐based filters and consistently give better pinpoint accuracy, especially in multi‐cluster cases. Training data consists of 140,000 SCF for ten landmark classes. The third stage, surface deformation, fits the facial template to the cranial based on the corresponding facial‐cranial landmarks. Five experts from the Anthropology department stated that of the reconstruction results, 91.5% could retain the template details and are accepted as the natural shape of the human face.
... The popular facial reconstruction methods, the American [3]. and Gerasimov [4], are the most intuitive method for a computer-based approach. Both simulate soft tissues guided by the placement of landmarks. ...
Cranial anthropometric reference points (landmarks) play an important role in craniofacial reconstruction and identification. Knowledge to detect the position of landmarks is critical. This work aims to locate landmarks automatically. Landmarks positioning using Surface Curvature Feature (SCF) is inspired by conventional methods of finding landmarks based on morphometrical features. Each cranial landmark has a unique shape. With the appropriate 3D descriptors, the computer can draw associations between shapes and landmarks using machine learning. The challenge in classification and detection in three-dimensional space is to determine the model and data representation. Using three-dimensional raw data in machine learning is a serious volumetric issue. This work uses the Surface Curvature Feature as a three-dimensional descriptor. It extracts the local surface curvature shape into a projection sequential value (depth). A machine learning method is developed to determine the position of landmarks based on local surface shape characteristics. Classification is carried out from the top-n prediction probabilities for each landmark class, from a set of predictions, then filtered to get pinpoint accuracy. The landmark prediction points are hypothetically clustered in a particular area, so a cluster-based filter is appropriate to isolate them. The learning model successfully detected the landmarks, with the average distance between the prediction points and the ground truth being 0.0326 normalized units. The cluster-based filter is implemented to increase accuracy compared to the ground truth. Thus, SCF is suitable as a 3D descriptor of cranial landmarks.
... Gerasimov suggested that the 'constitution of the subject', as a unique assemblage of facial features, played a more important role than 'race' in predicting a face based on the skull (Wilkinson, 2004, p. 122). Although Gerasimov (1971) repeatedly speaks of 'racial types', his aim was to go beyond these generalizations to create individual faces (Ullrich & Stephan, 2016). He argues that such stereotypical generalizations do not account for the variation within the perceived groups and therefore do not provide satisfying answers in the quest to predict individual faces. ...
The (re-)surfacing of race in forensic practices has received plenty of attention from STS scholars, especially in connection with modern forensic genetic technologies. In this article, I describe the making of facial depictions based on the skulls of unknown deceased individuals. Based on ethnographic research in the field of craniofacial identification and forensic art, I present a material-semiotic analysis of how race comes to matter in the face-making process. The analysis sheds light on how race as a translation device enables oscillation between the individual skull and population data, and allows for slippage between categories that otherwise do not neatly map on to one another. The subsuming logic of race is ingrained – in that it sits at the bases of standard choices and tools – in methods and technologies. However, the skull does not easily let itself be reduced to a racial type. Moreover, the careful efforts of practitioners to articulate the individual characteristics of each skull provide clues for how similarities and differences can be done without the effect of producing race. Such methods value the skull itself as an object of interest, rather than treat it as a vehicle for practicing race science. I argue that efforts to undo the persistence of race in forensic anthropology should focus critical attention on the socio-material configuration of methods and technologies, including data practices and reference standards.
... Arguably the most prolific reconstructionist was the Russian anthropologist Gerasimov who, a few years later, developed a method which combined previous approaches (gerassimoW 1968;uLLrich 1967;uLLrich & stephan 2016). Based on the cast of a skull he applied sculpted facial muscles and eyeballs, and finally modelled the skin surface. ...
In 1924 and 1925, anthropologist Egon von Eickstedt from the Natural History Museum of Vienna (NHMW), and Austrian/Hungarian artist Erna von Engel-Baiersdorf created two soft tissue reconstructions of the head of a Neanderthal, based on a cast of the skull from La Chapelle-aux-Saints, discovered in 1908. Eickstedt was to become a leading racial scientist and representative of German interwar and Nazi anthropology. Engel-Baiersdorf established herself as a scientific sculptor, survived the Holocaust, and reinvented herself as an anthropologist in Canada. The two busts were the first hominin reconstructions at the NHMW and initiated the NHMW's reconstruction workshop in the 1920s and 1930s. An original copy of the bust from 1924, which was recently rediscovered in the collection of the University Museum Utrecht, allows a detailed comparison with the 1925 bust in the NHMW collection in methodological terms: Eickstedt aimed at introducing a new method for facial reconstructions of fossil man, producing a 'racial type' or 'racial portrait', adopting and refining the reconstruction method developed by Kollmann & Büchly in 1898. A number of nineteenth and early twentieth century Western scientists discussed Neanderthals and modern Europeans in a triangular relationship with Indigenous peoples from German Pacific colonies. As we will show, the two early NHMW sculptures, as genuine products of German/Austrian interwar palaeo-raciology, combine theories and methods of ethnology , evolutionary and physical anthropology, and anatomy with artistic practices. Thus, they provide interesting new insights for current debates on the entanglements of German colonial history and the interwar/Nazi period.
... As the discipline of forensic facial approximation shifts towards an automated and computerised 3D approach, large databanks of population specific data must be created to create statistical models for predicting soft-tissue facial features based off the hard-tissue structures. Ear shape has been commonly ignored in the literature and standard casts used instead (Wilkinson 2010;Ullrich et al. 2016), but for an automatic facial approximation an accurate estimate of the ear based on ancestry, sex, and age specific guidelines is required. ...
The ear is a complex structure that has been mostly ignored in facial approximation. Thus, information is limited on how to reliably estimate its shape for craniofacial approximations. Current manual facial approximation methods employed in South Africa use standard ear casts, with no consideration of the influence of sex, age, and ancestry on the morphological structure of the ear. As the field of facial approximation moves towards automated and computerised methods, more studies are being aimed at developing sex-, age-, and ancestry-specific databases.
This study aims to assess variations and associations in ear shape and the eyes, nose, and mouth; and the underlying hard tissues of the external auditory meatus (EAM), nasal bones, nasal aperture, orbits, zygoma, and maxilla, in a sample of 40 black South Africans and 76 white South Africans between the ages of 18 and 90 years of age. A total of 50 capulometric and 43 craniometric landmarks were automatically placed on a sample of Cone Beam Computed Tomography (CBCT) scan reconstructions using MeVisLab 2.7.1. A further 559 semi-landmarks were automatically placed along the curves of the EAM, orbits, and anterior nasal aperture. The cartesian coordinates were recorded and analysed using geometric morphometric methods (GMM). General Procrustes Analysis (GPA), Principal Component Analysis (PCA), and multivariate normality testing were performed on all hard and soft tissue matrices for the entire sample, and each ancestral group separately.
Both hard- and soft-tissue auditory matrices resulted in statistically significant asymmetry (p-value = 0.007). Thus, left and right matrices for the ears and EAM were assessed individually. Statistical analysis performed using MANOVA, revealed highly significant variation in ear shape between groups for ancestry (p-value = 0.001), while sex and age were only significant between the white South African sub-sample (p -value < 0.05). The influence of ancestry in EAM shape was also found to be highly significant (p-value = 0.001), with sex only significantly influencing the right EAM and age not being significant. Size was only found to significantly influence shape on the auditory hard-tissues and not the soft-tissue ear. Strong positive correlations were observed between the soft-tissue ear and EAM (r2 > 0.7). The ear was also tested for correlations against other facial features, with strong positive correlations observed between shapes of the ear and orbit, mid-facial matrix, and nose – which is the facial feature most cited in the literature as the base of understanding for estimating the size and shape of the ear.
The results of this study support the use of automatic landmarking procedures to collect data on large samples, and the accurate placement of sliding landmarks will allow a better understanding of the shape of curves. Variations between groups indicate a need for population-specific databases for estimating the shape of the ear in South Africa. When estimating the shape of the ear, other facial features should be considered and factors such as the influence of ancestry included in the approximation. Sex and age will be of lesser concern when creating predictive models, as well as the influence of allometry.
... We will not be surprised if some readers support Gurche's method as found throughout the facial approximation literature is the view that that the face can be reliably approximated from the construction of the facial musculature alone [14,15,31]. However, as Ullrich and Stephan [58] have shown, this is a gross misinterpretation of the facial approximation method. In actual fact, facial approximation has always relied heavily on empirical data on soft tissue thicknesses [12,13,59]. ...
... Numerous facial approximation studies have developed methods for approximating the facial features in modern humans, although the validity of these methods applied to other hominids has never been tested. Published methods include the approximation of eyeball diameter and anatomical placement in the orbits [63]; eyebrow size, position and shape [64][65][66][67]; nasal profile [13,31,[68][69][70][71][72]; mouth width and shape [12,14,31,58,[73][74][75]; and size and shape of the external ear [11-13, 67, 76, 77]. Given that facial features are needed to complete any facial approximation, the interspecies compatibility of these methods is worthy of detailed examination in the near future to allow for complete approximations of Plio-Pleistocene hominids to be produced. ...
In modern humans, facial soft tissue thicknesses have been shown to covary with craniometric dimensions. However, to date it has not been confirmed whether these relationships are shared with non-human apes. In this study, we analyze these relationships in chimpanzees (Pan troglodytes) with the aim of producing regression models for approximating facial soft tissue thicknesses in Plio-Pleistocene hominids. Using CT scans of 19 subjects, 637 soft tissue, and 349 craniometric measurements, statistically significant multiple regression models were established for 26 points on the face and head. Examination of regression model validity resulted in minimal differences between observed and predicted soft tissue thickness values. Assessment of interspecies compatibility using a bonobo (Pan paniscus) and modern human subject resulted in minimal differences for the bonobo but large differences for the modern human. These results clearly show that (1) soft tissue thicknesses covary with craniometric dimensions in P. troglodytes, (2) confirms that such covariation is uniformly present in both extant Homo and Pan species, and (3) suggests that chimp-derived regression models have interspecies compatibility with hominids who have similar craniometric dimensions to P. troglodytes. As the craniometric dimensions of early hominids, such as South African australopithecines, are more similar to P. troglodytes than those of H. sapiens, chimpanzee-derived regression models may be used for approximating their craniofacial anatomy. It is hoped that the results of the present study and the reference dataset for facial soft tissue thicknesses of chimpanzees it provides will encourage further research into this topic.
