Facial soft tissue depth of a contemporary adult Greek population
- Open Access
- 10.08.2024
- Original Article
Erschienen in:
Abstract
Introduction
The identification of deceased individuals has legal and ethical significance in a medico-legal context [1]. Establishing the identity of unknown individuals facilitates solving criminal cases (e.g., homicide and unexpected natural death), identifying the victims of natural disasters (e.g., earthquakes, hurricanes, tsunamis) and war crimes (e.g., genocide) [2‐4]. From an ethical point of view, returning bodies to their loved ones mitigates the grief of families as well as concludes the legal issues such as the death certificate, inheritance, life insurance benefit and so on [5, 6].
Identification of skeletonised remains is achieved by one of the positive identification methods, such as deoxyribonucleic acid (DNA) analysis and the comparison of dental records depending on the preservation condition of the remains and the availability of the antemortem data of the person in question [4]. Yet, these positive identification methods are not feasible if there is no antemortem data available [4, 7]. In this event, presumptive or possible identification methods are used to narrow down possible identities by eliminating mismatching data. The presumptive identification method consists of the anthropological examination of skeletal remains and requires establishing the biological profile of an unknown individual [8]. This profiling includes age, sex, stature, ancestral background, skeletal variation/ abnormalities and congenital or pathology/trauma-related anomalies on bones [2, 9‐13].
Anzeige
When positive and presumptive identification methods are insufficient in pointing an identity, the facial approximation of an unidentified individual is performed as a last resort to create a face that resembles the appearance of the deceased in life to reach individuals who might know the possible identity [14]. The method relies on the application of facial soft tissue depth (FSTD) and the creation of facial features (eyes, nose, mouth, ears, etc.) [13]. To date, several population-specific FSTD datasets have been set to create precise depth datasets for facial approximation [15‐26]. People of Greek ancestry are one of the population groups for which a FSTD dataset has not been generated yet. Therefore, the aim of this study is to fill the gap in the field and establish a FSTD database for adult Greeks.
Materials and methods
The material of this study consisted of the head CT scans of 100 subjects (50 male and 50 female), with an average age of 59 years old (ranging between 18 and 99 years). The subjects underwent CT scans for diagnostic purposes at the radiology department of Athens Medical Centre, Athens, Greece. Thus, the individuals were not exposed to radiation specifically for this study. All the subjects were assumed Greek based on their surnames. Any self-confirmation of ancestral background or genetic evidence was not collected.
CT scan protocol, landmarks and FSTD measurement
The CT scans were taken using the Siemens SOMATOM® Sensation 64 (Siemens AG, Forchheim, Germany). The images were recorded following the tube current of 400 mAs, at the tube voltage of 20 kV, with a slice thickness of 2.4 mm and a matrix of 512 × 512 pixels. The subjects were scanned when they were in the supine position. All the images were obtained in the DICOM format (Digital Imaging and Communications in Medicine). Along with the CT scans, the age and sex information of the subjects were obtained for the statistical analyses.
Commonly used 5 mid-sagittal and 17 bilateral cranial landmarks (Fig. 1; Table 1) were selected for this study to generate a comprehensive FSTD database. However, only those landmarks found on the upper portion of the skull were included in this study since the radiological examination did not require scanning the whole head, only the upper two-thirds of the skulls were scanned (from the top of the skull to the mastoid processes). Although the nasomaxillofrontale was identified on each skull to assist in locating the mid-nasomaxillare, FSTD from this landmark was not collected due to the obstruction caused by the eyelids in most of the cases.
Anzeige
Before measuring the facial depths, the skulls and heads of the individuals were segmented using the histogram method [29] in AMIRA 6.1 (Thermo Fischer Scientific, Cleveland, OH, USA). Prior to taking measurements, the 3D skull models were positioned in the Frankfurt Horizontal plane (FHP), which implies that a plane passes through both right and left porions and both right and left orbitales [30]. When the skulls were still in the FHP, each cranial landmark was placed on each skull individually by following the description given in Table 1. This process was carried out in the sagittal, coronal and sagitto-coronal planes depending on the location of the landmarks (Table 1), as described by Guyomarc’h et al. [17]. Once all the cranial landmarks were placed, the 3D head and skull models were superimposed using the surface generation algorithm of Amira (Fig. 2). This algorithm provides visualising surfaces in different styles (e.g., solid, transparent, etc.). The head models were created using the transparent surface feature to ensure that the cranial landmarks could be easily identified and traced with precision (Fig. 2). The cephalometric landmarks were then placed on the 3D head models by perpendicularly tracing the cranial landmarks. Subsequently, the horizontal distance between the cranial and cephalometric landmarks was measured. All the measurements were taken by the first author.
Table 1
The definitions, abbreviations and synonyms of the cranial landmarks selected for the current study
Cranial Landmarks | Definition | Synonym |
|---|---|---|
1. Supraglabella (Sg) | On the frontal bone, 10 mm superior to the glabella20 | Ophyron20 |
2. Glabella (G) | Most anterior point between supraorbital ridges18 | |
3. Nasion (N) | The midpoint of the suture between the frontal and the two nasal bones18 | |
4. Mid-nasal (Mn) | On the internasal suture, midway between the nasion and rhinion20 | |
5. Rhinion (Rhi) | The anterior tip of the nasal bone16 | End of nasals17 |
6. Frontal Eminence (Fe) | Place on the projections at both sides of the forehead15 | |
7. Supraorbital (So) | Most anterior point of the supraciliary arch in the axe of the centre of the orbit20 | Superciliare17; superior eye orbit17 Mid-supraorbital20 |
8. Supraconchion (Sk) | Most superior point of the orbital rim17 | Orbitale superius17 |
9. Orbitale (Or) | Most inferior point of the orbital margin17 | Inferior eye orbit17 |
10. Sub-maxillar Curvature (Smc) | Most supero-medial point on the maxillary inflexion between the zygomaxillare and the ectomolare17 | |
11. Frontotemporale (Ft) | Most anterio-medial point of the linea temporalis superior17 | |
12. Frontomalare Temporale (Fmt) | Most posterior point of the zygo-frontal suture17 | Mid-lateral orbit17; lateral eye orbit17 |
13. Ectoconchion (Ec) | The most lateral point at the lateral margin of orbit19 | |
14. Mid-zygomatic (Mz) | Lined up with the lateral border of the orbit on the centre of the zygomatic process20 | Lateral orbit20; zygomatic malare20 |
15. Zygomaxillare (Zm) | Most inferior point on the zygo-maxillary suture17 | |
16. Jugale (Ju) | Most antero-inferior point on the posterior border of the zygomatic bone17 | |
17. Zygion (Zy) | Most lateral extent of the lateral surface of the zygomatic arch17 | Root of zygoma17; Lateral zygomatic arch17; Zygomatic arch27 |
18. Condylion (Co) | Most lateral point of the glenoid process of the mandible20 | Condylion laterale20 |
19. Alare (Al) | Most lateral point on the margin of the anterior nasal aperture19 | Apertion20 |
20.Mid-nasomaxillare (Mnm) | Mid-point of the naso-maxillary suture between the nasomaxillare and the nasomaxillofrontale17 | Lateral nasal17 |
21. Nasomaxillare (Nm) | Most inferior point of the naxo-maxillary suture on the nasal aperture17 | |
22. Frontomalare Orbitale (Fmo) | Point where the frontozygomatic suture crosses the inner orbital rim20 | |
23. Nasomaxillofrontale (Nmf) | Junction of the frontal, maxillary and lacrimal bones on the medial bone of the orbit20 | Lateral Glabella20 |
Fig. 1
Anterior (A) and lateral (B) views of the 3D skull model, showing the cranial landmarks. (*): Nasomaxillofrontale. Due to the slice thickness of the CT scans (2.4 mm), horizontal lines appeared on the skull models
Accuracy assessment
Inaccurate segmentation of the 3D skull and head models might impact the accuracy of FSTD measurements. Therefore, before proceeding to the facial depth measurement process, the reproducibility and precision of the 3D models were tested by re-segmenting the head and skull models of randomly selected eight subjects at the five-week interval. The results were evaluated using Meshlab (Visual Computing Lab-ISTI-CNR, Pisa& Rome, Italy) by calculating the Hausdorff distance (HD), which is the longest distance between the two closest points on each 3D model pair when they are superimposed [31]. When HD equals zero, it indicates that the two models are identical. On the other hand, the HD values above 0.5 mm are considered a poor re-segmentation. Thus, the HD values that are close to zero suggest an accurate segmentation [35].
Additionally, the accuracy, repeatability and reproducibility of the measurements were calculated for intra- and inter-observer errors. The former was determined by the first author by re-measuring the FSTD of randomly selected subjects (n = 25) at a four-week interval. The latter, however, was measured by a second examiner who has no prior experience in identifying the cranial landmarks on 3D skull models and measuring FSTD. These measurements were re-collected on randomly selected 12 subjects at a four-week interval by following the same measurement protocol. The differences between the repeating measurements were determined by calculating the technical error of measurements (TEM) by applying the following formula:
$$\text{TEM}=\sqrt{\frac{{D}2}{2{N}}}$$
Where D is the difference between the first and the second measurements, N is the number of measurements re-taken [32, 33]. While the small TEM values are associated with the low TEM, the large mean values are an indication of the high TEM values. This positive correlation between the TEM values and the size of the measurements causes difficulty when two different sizes of measurements or variables are compared [34]. In order to tackle this problem, the TEM value is converted into the relative technical error measurements (rTEM) so as to indicate the error margin in percentage. This conversion is made by the formula given:
$$\text{r}\,\text{TEM}=\frac{{TEM}}{{VAV}}\times100$$
Anzeige
Where VAV is the variable average value, and it was calculated by computing the arithmetic mean value of the mean values of the first and second measurements [33]. The results lower than 5% are considered reliable [32, 34]. Lastly, the coefficient of reliability (R) was calculated to determine the portion of error-free measurements by applying the formula given below:
$$\text{R}=1-[\frac{{\left(TEM\right)}^{2}}{{S}^{2}}]$$
Statistical analysis
Data were analyzed using the Statistical Package for Social Sciences (SPSS, version 25.0). General descriptive analysis was performed for each landmark and mean facial soft tissue depths were calculated considering the age and sex categories of the individuals. Prior to performing statistical analyses, the individuals were divided into five age categories as follows: 18–34 years, 35–44 years, 45–54 years, 55–64 years and ≥ 65 years.
Anzeige
The significance level of < 0.05 was set for all the statistical tests. The distribution of normality was tested by using the Shapiro-Wilk test. The significance of the impact of age and sex on FSTD was tested by performing a multivariate analysis of variance (MANOVA). It should be noted that the impact of body mass index (BMI) on FSTD couldn’t be measured since the height and weight information of patients were not recorded at the hospital. A paired sample t-test was used to quantify the facial asymmetry for normally distributed data and the Wilcoxon-signed rank test for non-normally distributed data. The FSTDs of this study were compared with the reported mean FSTD values of five populations [17, 19, 22, 23, 37] which have a similar measurement protocol. The comparisons were performed by using the one sample t-test for normally distributed data and the Wilcoxon one-sample signed-ranked test was utilised for non-normally distributed data. Bootstrap 1000 was used for all statistical tests in order to explain possible biases that occurred due to the small sample size.
Fig. 2
The superimposed 3D skull and head models in the sagittal plane
Results
Accuracy assessment of skull and head segmentations
The results showed that the head and skull models of this study were reproduced with good precision. All the HD values of the head models were lower than 0.5 mm, ranging between 0.01 and 0.15 mm (Table 2). The lowest HD value (high precision) was recorded from Head 1. The same HD values were reported from Heads 5, 6 and 8, which was 0.02 mm. The highest HD score (low precision) was reported from Head 3 with 0.15 mm, which was still significantly below 0.5 mm [35]. The HD values of the skull models were slightly higher than those of the head models (Table 2). The only HD value that exceeded 0.5 mm was the Skull 4 with 0.63 mm.
Table 2
Hausdorff distances (HD) calculated for the skull and head models (measurements are in mm)
Skull/Head No | Head | Skull |
|---|---|---|
Hausdorff distance | Hausdorff distance | |
1 | 0.01 | 0.04 |
2 | 0.12 | 0.36 |
3 | 0.15 | 0.36 |
4 | 0.11 | 0.63 |
5 | 0.02 | 0.14 |
6 | 0.02 | 0.08 |
7 | 0.08 | 0.01 |
8 | 0.02 | 0.03 |
Anzeige
HD values suggested that the head models were re-segmented with higher precision in comparison with the skull re-segmentations. This difference was due to the contrast between neighbouring voxels. When segmenting the head models, the contrast between the voxels of air (black) and the head (grey) was high. Therefore, the colour difference between the voxels was more distinguishable. On the other hand, differentiating the voxels belonging to the skull (light grey/white) from those of the head (grey) was rather challenging due to the partial volume effect [36]. Hence, the skull re-segmentations were completed with less precision compared to the head models.
Intra- and inter-observer error
The results of the intra-observer error assessment showed that all the measurements were within the acceptable accuracy range. The absolute TEM values were generally low, fluctuating between 0.10 mm and 1.05 mm (Table 3). The highest FSTD values were generally collected from the cheek region. Since there is a positive correlation between the high-depth measurements and high TEM values, obtaining high TEM values from the cheek region was anticipated. In order to solve this, the TEM values were converted into the rTEM values. The rTEM values were at or below 5%, ranging between 1.6% and 5%. The highest rTEM value was obtained from the mid-nasal (5%), yet it was still considered a repeatable measurement [32]. The R values were, however, close to 1, ranging between 0.84 and 0.99 mm. The lowest R values were obtained for the left and right sub-maxillar curvature (0.88 and 0.84 mm, respectively). However, they were still within the acceptable accuracy range [32].
The absolute TEM values calculated for the inter-observer error varied between 0.19 and 1.45 mm. The high absolute TEM values were calculated for the left zygomaxillare, left and right sub-maxillar curvature, right condylion and right nasomaxillare. The rTEM values showed significant variation compared to those of the intra-observer error. The rTEM percentages of 22 landmarks were above 5% (Table 3). Therefore, the repeated measurements of the 22 landmarks were considered unreliable for the inter-observer error. On the other hand, the rTEM scores of the 17 landmarks were found to be reliable. The R values of inter-observer error were mostly within the 0.80–1 mm accuracy range. Only the mid-nasal, right sub-maxillar curvature, left zygomaxillare and right nasomaxillare were lower than 0.80 mm. These results suggested that experience impacts the accuracy of FSTD measurements.
Table 3
TEM, rTEM and coefficient of reliability (R) results reveal intra- and inter-observer errors calculated for the 3D measurement method
Intra-Observer Error | Inter-Observer Error | ||||||||
|---|---|---|---|---|---|---|---|---|---|
Landmarks | n | TEM (mm) | rTEM (%) | R | n | TEM (mm) | rTEM (%) | R | |
Sg | 25 | 0.20 | 3.5 | 0.97 | 12 | 0.27 | 4.53 | 0.95 | |
G | 25 | 0.10 | 1.8 | 0.99 | 12 | 0.20 | 3.18 | 0.96 | |
N | 25 | 0.11 | 1.6 | 0.99 | 12 | 0.23 | 3.48 | 0.96 | |
Mn | 25 | 0.14 | 5.0 | 0.97 | 12 | 0.47 | 18.35 | 0.63 | |
Rhi | 25 | 0.16 | 4.7 | 0.94 | 12 | 0.23 | 6.64 | 0.93 | |
FeL | 25 | 0.21 | 3.3 | 0.97 | 12 | 0.43 | 6.36 | 0.94 | |
FeR | 25 | 0.21 | 3.3 | 0.97 | 12 | 0.66 | 9.47 | 0.93 | |
SoL | 25 | 0.26 | 3.0 | 0.97 | 12 | 0.38 | 5.00 | 0.95 | |
SoR | 25 | 0.17 | 1.9 | 0.98 | 12 | 0.48 | 6.05 | 0.93 | |
SkL | 25 | 0.29 | 3.2 | 0.98 | 12 | 0.89 | 9.76 | 0.84 | |
SkR | 25 | 0.37 | 4.2 | 0.95 | 12 | 0.47 | 5.07 | 0.91 | |
OrL | 25 | 0.14 | 2.1 | 0.99 | 12 | 0.32 | 4.65 | 0.98 | |
OrR | 25 | 0.20 | 2.9 | 0.99 | 12 | 0.33 | 4.86 | 0.97 | |
SmcL | 25 | 0.94 | 3.4 | 0.88 | 12 | 1.08 | 3.98 | 0.80 | |
SmcR | 24 | 1.05 | 3.8 | 0.84 | 12 | 1.06 | 3.78 | 0.66 | |
FtL | 25 | 0.29 | 4.5 | 0.96 | 12 | 0.40 | 6.60 | 0.92 | |
FtR | 25 | 0.27 | 4.2 | 0.98 | 12 | 0.48 | 6.73 | 0.95 | |
FmtL | 25 | 0.22 | 3.2 | 0.98 | 12 | 0.61 | 9.70 | 0.80 | |
FmtR | 25 | 0.32 | 3.8 | 0.97 | 12 | 0.64 | 9.47 | 0.89 | |
EcL | 25 | 0.26 | 2.9 | 0.99 | 12 | 0.81 | 9.67 | 0.91 | |
EcR | 25 | 0.22 | 2.6 | 0.99 | 12 | 0.96 | 11.39 | 0.86 | |
MzL | 25 | 0.15 | 1.6 | 0.99 | 12 | 0.21 | 2.18 | 0.98 | |
MzR | 25 | 0.23 | 2.2 | 0.98 | 12 | 0.30 | 2.97 | 0.97 | |
ZmL | 25 | 0.64 | 4.5 | 0.90 | 12 | 1.45 | 8.69 | 0.58 | |
ZmR | 25 | 0.73 | 4.7 | 0.90 | 12 | 0.85 | 4.98 | 0.84 | |
JuL | 25 | 0.48 | 4.8 | 0.93 | 12 | 0.26 | 2.62 | 0.98 | |
JuR | 25 | 0.32 | 2.9 | 0.96 | 12 | 0.44 | 4.15 | 0.94 | |
ZyL | 25 | 0.30 | 3.1 | 0.98 | 12 | 0.22 | 2.22 | 0.99 | |
ZyR | 25 | 0.22 | 2.3 | 0.99 | 12 | 0.23 | 2.36 | 0.99 | |
CoL | 25 | 0.41 | 2.2 | 0.98 | 12 | 0.19 | 0.97 | 0.99 | |
CoR | 25 | 0.40 | 2.2 | 0.98 | 12 | 1.12 | 5.88 | 0.93 | |
AlL | 25 | 0.27 | 2.8 | 0.98 | 12 | 0.59 | 5.93 | 0.95 | |
AlR | 25 | 0.30 | 3.1 | 0.98 | 12 | 0.42 | 4.62 | 0.96 | |
MnmL | 25 | 0.15 | 4.8 | 0.98 | 12 | 0.31 | 10.88 | 0.94 | |
MnmR | 24 | 0.15 | 4.8 | 0.98 | 12 | 0.44 | 13.91 | 0.90 | |
NmL | 25 | 0.25 | 4.5 | 0.98 | 12 | 0.78 | 14.86 | 0.84 | |
NmR | 25 | 0.25 | 4.2 | 0.98 | 12 | 1.21 | 22.94 | 0.73 | |
FmoL | 25 | 0.24 | 2.5 | 0.99 | 12 | 0.85 | 9.97 | 0.93 | |
FmoR | 25 | 0.38 | 4.0 | 0.98 | 12 | 0.95 | 11.56 | 0.91 | |
FSTD dataset
Tables 4, 5, 6, 7 and 8 show the descriptive statistics for the FSTDs measured at each anatomic landmark, considering the age groups (18–34, 35–44, 45–54, 55–64 and ≥ 65 years) of both males and females.
Table 4
Facial soft tissue thicknesses of the male and female individuals that were obtained using the 3D measurement method for the 18–34 years age group (measurements in mm)
Landmarks | Male (n = 8) | Female (n = 8) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
n | \(\:\stackrel{-}{\varvec{\upchi\:}}\) | σ | Min | Max | n | \(\:\stackrel{-}{\varvec{\upchi\:}}\) | σ | Min | Max | |
Sg | 8 | 6.48 | 2.52 | 4.70 | 8.62 | 7 | 4.79 | 0.72 | 4.00 | 6.14 |
G | 8 | 5.99 | 2.02 | 4.56 | 7.73 | 8 | 4.85 | 0.56 | 4.20 | 8.69 |
N | 8 | 7.30 | 0.54 | 5.80 | 9.65 | 8 | 5.57 | 1.08 | 4.36 | 9.98 |
Mn | 5 | 4.17 | 1.07 | 2.02 | 4.93 | 7 | 2.39 | 0.71 | 1.55 | 3.33 |
Rhi | 6 | 3.09 | 0.17 | 2.09 | 4.63 | 8 | 2.57 | 0.64 | 1.97 | 3.24 |
FeL | 6 | 6.48 | 0.48 | 6.14 | 6.97 | 7 | 5.93 | 0.90 | 4.85 | 6.95 |
FeR | 6 | 5.96 | 1.22 | 5.10 | 7.15 | 7 | 6.57 | 1.89 | 4.39 | 7.76 |
SoL | 8 | 8.25 | 0.28 | 7.69 | 11.53 | 8 | 6.77 | 0.66 | 6.01 | 12.64 |
SoR | 8 | 8.11 | 0.25 | 7.35 | 10.90 | 8 | 6.56 | 1.30 | 5.14 | 10.65 |
SkL | 8 | 9.54 | 3.01 | 7.41 | 11.67 | 8 | 6.66 | 0.62 | 5.97 | 12.34 |
SkR | 8 | 8.78 | 2.26 | 7.18 | 10.64 | 8 | 6.05 | 1.00 | 4.99 | 11.75 |
OrL | 8 | 5.74 | 3.21 | 3.47 | 12.04 | 8 | 4.87 | 2.28 | 3.31 | 8.64 |
OrR | 8 | 6.44 | 3.55 | 3.93 | 9.84 | 8 | 5.16 | 1.79 | 3.95 | 8.12 |
SmcL | 7 | 22.71 | 2.11 | 21.22 | 35.71 | 6 | 23.63 | 2.63 | 21.24 | 28.93 |
SmcR | 6 | 25.54 | 6.00 | 21.30 | 34.00 | 6 | 25.94 | 2.24 | 23.78 | 29.82 |
FtL | 8 | 5.76 | 0.58 | 4.43 | 9.97 | 8 | 5.09 | 0.80 | 4.37 | 7.33 |
FtR | 8 | 5.94 | 0.73 | 4.26 | 9.11 | 8 | 4.99 | 0.35 | 4.64 | 8.37 |
FmtL | 8 | 7.70 | 2.07 | 5.32 | 10.24 | 8 | 5.35 | 0.93 | 4.30 | 9.31 |
FmtR | 8 | 9.92 | 0.95 | 6.16 | 11.25 | 8 | 6.70 | 2.34 | 4.06 | 10.00 |
EcL | 8 | 6.48 | 0.01 | 4.11 | 14.90 | 8 | 10.85 | 6.72 | 3.64 | 16.95 |
EcR | 6 | 5.64 | 0.70 | 5.14 | 16.22 | 8 | 7.72 | 4.72 | 2.32 | 11.06 |
MzL | 7 | 7.14 | 0.83 | 6.21 | 13.25 | 8 | 7.34 | 0.96 | 6.26 | 11.08 |
MzR | 7 | 7.91 | 0.59 | 7.49 | 12.18 | 8 | 8.37 | 0.76 | 7.89 | 12.27 |
ZmL | 7 | 12.07 | 0.54 | 11.03 | 20.07 | 8 | 11.46 | 2.36 | 9.83 | 17.75 |
ZmR | 7 | 13.90 | 1.83 | 12.60 | 17.83 | 8 | 12.40 | 2.39 | 10.94 | 19.60 |
JuL | 7 | 8.63 | 0.51 | 7.06 | 13.81 | 8 | 7.79 | 1.26 | 7.04 | 12.64 |
JuR | 7 | 9.11 | 0.22 | 7.73 | 12.27 | 8 | 8.83 | 1.34 | 7.65 | 11.51 |
ZyL | 8 | 7.55 | 0.13 | 5.16 | 13.97 | 8 | 7.86 | 0.98 | 6.93 | 12.43 |
ZyR | 7 | 7.49 | 0.74 | 5.34 | 12.42 | 8 | 8.11 | 0.76 | 7.32 | 13.01 |
CoL | 8 | 16.77 | 1.17 | 13.46 | 29.90 | 8 | 13.70 | 2.27 | 11.88 | 21.38 |
CoR | 8 | 16.15 | 2.34 | 13.14 | 28.86 | 8 | 12.98 | 1.48 | 11.27 | 20.37 |
AlL | 7 | 9.03 | 1.88 | 7.70 | 14.52 | 7 | 7.68 | 0.88 | 5.92 | 12.25 |
AlR | 7 | 10.44 | 1.51 | 7.48 | 12.27 | 7 | 7.94 | 1.49 | 6.09 | 13.10 |
MnmL | 6 | 2.57 | 0.01 | 2.56 | 4.79 | 8 | 2.52 | 0.31 | 2.21 | 5.00 |
MnmR | 6 | 2.60 | 0.22 | 2.44 | 4.81 | 8 | 2.52 | 0.56 | 2.05 | 4.76 |
NmL | 6 | 4.39 | 0.26 | 4.20 | 9.46 | 7 | 4.16 | 0.36 | 3.52 | 5.99 |
NmR | 6 | 5.07 | 1.13 | 4.27 | 9.32 | 5 | 4.65 | 0.36 | 2.95 | 5.04 |
FmoL | 8 | 6.89 | 1.57 | 5.78 | 14.25 | 8 | 10.15 | 2.65 | 6.94 | 14.49 |
FmoR | 8 | 6.16 | 0.21 | 6.01 | 14.64 | 8 | 8.92 | 2.71 | 5.61 | 14.13 |
Table 5
Facial soft tissue thicknesses of the male and female individuals that were obtained using the 3D measurement method for the 35–44 years age group (measurements in mm)
Male (n = 5) | Female (n = 5) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Landmarks | n | \(\:\stackrel{-}{\varvec{\upchi\:}}\) | σ | Min | Max | n | \(\:\stackrel{-}{\varvec{\upchi\:}}\) | σ | Min | Max |
Sg | 5 | 7.62 | 1.41 | 4.00 | 8.61 | 5 | 5.49 | 1.65 | 4.32 | 6.66 |
G | 5 | 6.89 | 2.24 | 4.37 | 8.97 | 5 | 5.64 | 1.41 | 4.64 | 6.67 |
N | 5 | 8.79 | 1.18 | 5.51 | 10.95 | 5 | 6.11 | 2.09 | 4.63 | 7.98 |
Mn | 5 | 3.37 | 1.21 | 2.46 | 6.67 | 5 | 2.80 | 0.65 | 1.80 | 3.26 |
Rhi | 5 | 4.22 | 0.59 | 2.40 | 4.63 | 5 | 2.47 | 0.59 | 2.05 | 3.05 |
FeL | 5 | 7.64 | 0.83 | 4.60 | 8.49 | 5 | 6.61 | 2.06 | 5.15 | 8.06 |
FeR | 5 | 6.60 | 0.43 | 4.18 | 9.16 | 5 | 5.49 | 1.32 | 4.56 | 7.42 |
SoL | 5 | 11.48 | 1.44 | 9.16 | 12.50 | 5 | 7.22 | 0.25 | 6.37 | 9.24 |
SoR | 5 | 10.94 | 0.37 | 9.71 | 11.70 | 5 | 7.07 | 0.93 | 6.41 | 8.60 |
SkL | 5 | 10.92 | 2.26 | 8.87 | 14.53 | 5 | 6.92 | 0.71 | 6.41 | 8.25 |
SkR | 4 | 10.74 | 1.39 | 9.59 | 11.72 | 5 | 6.54 | 0.69 | 6.05 | 8.12 |
OrL | 5 | 9.19 | 3.15 | 3.96 | 11.42 | 5 | 5.55 | 0.08 | 5.20 | 6.83 |
OrR | 5 | 8.97 | 2.88 | 4.20 | 11.01 | 5 | 4.91 | 1.73 | 3.68 | 7.47 |
SmcL | 3 | 27.67 | 1.79 | 23.47 | 28.93 | 4 | 27.43 | 3.68 | 24.83 | 30.03 |
SmcR | 3 | 29.77 | 4.70 | 26.44 | 33.09 | 4 | 27.81 | 4.56 | 24.58 | 31.03 |
FtL | 5 | 7.55 | 1.67 | 3.43 | 8.73 | 5 | 5.28 | 1.85 | 3.97 | 6.91 |
FtR | 5 | 7.79 | 1.48 | 3.10 | 8.83 | 5 | 4.85 | 0.57 | 3.97 | 7.99 |
FmtL | 5 | 8.42 | 1.00 | 7.29 | 9.13 | 5 | 5.82 | 2.04 | 3.60 | 10.03 |
FmtR | 5 | 8.67 | 2.35 | 6.52 | 10.33 | 5 | 6.06 | 2.89 | 4.01 | 11.52 |
EcL | 5 | 11.47 | 1.13 | 3.26 | 14.40 | 5 | 6.50 | 1.36 | 4.49 | 12.67 |
EcR | 4 | 9.33 | 4.24 | 4.55 | 12.32 | 5 | 6.21 | 1.70 | 2.32 | 12.60 |
MzL | 3 | 9.57 | 2.31 | 7.93 | 11.20 | 5 | 9.88 | 1.53 | 8.80 | 11.98 |
MzR | 3 | 10.27 | 1.24 | 9.39 | 11.33 | 5 | 9.71 | 1.32 | 8.77 | 13.32 |
ZmL | 3 | 14.93 | 0.23 | 14.76 | 15.85 | 5 | 13.34 | 2.18 | 11.79 | 16.35 |
ZmR | 3 | 16.96 | 0.27 | 16.77 | 17.57 | 5 | 13.60 | 1.47 | 12.56 | 21.20 |
JuL | 5 | 9.29 | 0.44 | 7.76 | 9.62 | 5 | 8.62 | 3.37 | 6.23 | 12.96 |
JuR | 5 | 9.86 | 1.25 | 7.45 | 10.74 | 5 | 9.55 | 3.22 | 7.27 | 13.22 |
ZyL | 5 | 9.89 | 0.70 | 7.60 | 10.38 | 5 | 9.54 | 5.91 | 5.36 | 13.72 |
ZyR | 5 | 10.85 | 1.46 | 6.93 | 11.88 | 5 | 9.64 | 5.15 | 6.00 | 13.28 |
CoL | 5 | 20.13 | 3.99 | 16.57 | 23.01 | 5 | 13.98 | 7.61 | 8.60 | 25.82 |
CoR | 5 | 19.38 | 4.84 | 13.82 | 25.71 | 5 | 13.22 | 6.00 | 8.98 | 25.17 |
AlL | 3 | 10.52 | 1.08 | 8.90 | 11.28 | 5 | 8.60 | 0.38 | 7.04 | 13.03 |
AlR | 3 | 9.04 | 0.49 | 8.69 | 9.84 | 5 | 8.00 | 1.79 | 6.73 | 11.56 |
MnmL | 5 | 3.56 | 0.59 | 2.50 | 3.98 | 5 | 2.43 | 0.04 | 1.81 | 4.89 |
MnmR | 5 | 4.01 | 1.42 | 2.41 | 5.01 | 5 | 2.33 | 0.13 | 2.23 | 4.94 |
NmL | 5 | 5.88 | 0.35 | 3.74 | 9.18 | 4 | 4.59 | 0.18 | 4.33 | 5.65 |
NmR | 5 | 6.69 | 1.22 | 4.65 | 8.39 | 3 | 4.49 | 0.46 | 4.16 | 4.83 |
FmoL | 5 | 15.80 | 1.18 | 4.98 | 16.63 | 5 | 9.59 | 2.28 | 5.22 | 11.41 |
FmoR | 5 | 14.74 | 1.07 | 6.39 | 17.87 | 5 | 8.33 | 1.83 | 5.30 | 11.29 |
Table 6
Facial soft tissue thicknesses of the male and female individuals that were obtained using the 3D measurement method for the 45–54 years age group (measurements in mm)
Landmarks | Male (n = 9) | Female (n = 6) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
n | \(\:\stackrel{-}{\varvec{\upchi\:}}\) | σ | Min | Max | n | \(\:\stackrel{-}{\varvec{\upchi\:}}\) | σ | Min | Max | |
Sg | 9 | 5.19 | 0.46 | 4.74 | 8.21 | 6 | 6.01 | 0.74 | 5.49 | 7.75 |
G | 9 | 5.35 | 0.39 | 4.29 | 8.70 | 6 | 6.01 | 0.78 | 5.46 | 7.69 |
N | 9 | 6.95 | 1.09 | 4.98 | 10.71 | 6 | 7.28 | 1.65 | 6.11 | 8.45 |
Mn | 8 | 3.14 | 0.70 | 1.90 | 4.13 | 5 | 2.98 | 1.05 | 2.24 | 3.74 |
Rhi | 8 | 3.53 | 0.77 | 2.32 | 4.57 | 5 | 2.91 | 0.40 | 2.04 | 3.27 |
FeL | 8 | 5.97 | 1.52 | 4.28 | 7.52 | 6 | 5.48 | 1.36 | 4.52 | 6.74 |
FeR | 9 | 5.89 | 1.54 | 4.11 | 8.24 | 6 | 5.51 | 1.80 | 4.23 | 7.80 |
SoL | 8 | 9.25 | 0.26 | 6.94 | 12.19 | 6 | 8.33 | 0.69 | 7.10 | 10.25 |
SoR | 8 | 8.81 | 0.76 | 7.70 | 12.91 | 6 | 8.20 | 1.12 | 7.41 | 9.58 |
SkL | 9 | 9.83 | 0.85 | 6.06 | 14.20 | 6 | 8.51 | 0.71 | 7.58 | 10.02 |
SkR | 8 | 10.02 | 0.97 | 7.64 | 13.50 | 6 | 9.15 | 0.64 | 8.27 | 9.93 |
OrL | 9 | 8.37 | 2.57 | 2.65 | 12.08 | 6 | 5.52 | 0.12 | 3.57 | 7.84 |
OrR | 9 | 7.62 | 2.59 | 3.33 | 11.55 | 6 | 5.67 | 1.00 | 3.57 | 9.54 |
SmcL | 7 | 27.73 | 0.73 | 26.90 | 36.07 | 5 | 25.62 | 0.62 | 25.18 | 31.31 |
SmcR | 7 | 29.11 | 1.51 | 27.59 | 35.24 | 5 | 26.36 | 0.52 | 24.99 | 29.16 |
FtL | 9 | 5.78 | 1.41 | 4.78 | 8.84 | 6 | 6.42 | 1.03 | 4.01 | 9.59 |
FtR | 9 | 5.49 | 1.22 | 4.13 | 7.65 | 6 | 6.76 | 1.54 | 3.90 | 8.59 |
FmtL | 9 | 8.82 | 1.61 | 6.58 | 10.51 | 6 | 6.69 | 3.64 | 4.11 | 9.26 |
FmtR | 9 | 8.62 | 1.54 | 6.51 | 11.92 | 6 | 7.40 | 3.94 | 4.61 | 10.41 |
EcL | 9 | 13.48 | 2.29 | 5.56 | 19.16 | 6 | 10.03 | 4.38 | 5.63 | 13.12 |
EcR | 9 | 12.19 | 0.97 | 6.21 | 16.44 | 6 | 9.08 | 3.19 | 5.91 | 11.33 |
MzL | 6 | 8.02 | 1.14 | 6.92 | 9.99 | 5 | 8.78 | 0.47 | 8.44 | 12.05 |
MzR | 6 | 8.84 | 1.67 | 7.29 | 10.61 | 5 | 9.19 | 0.74 | 8.66 | 12.43 |
ZmL | 6 | 13.26 | 1.18 | 11.90 | 18.44 | 5 | 13.00 | 0.86 | 12.39 | 16.85 |
ZmR | 6 | 15.78 | 0.84 | 14.81 | 17.26 | 5 | 14.08 | 0.13 | 13.98 | 19.33 |
JuL | 9 | 10.45 | 0.98 | 7.95 | 13.81 | 6 | 9.88 | 0.18 | 9.75 | 14.19 |
JuR | 9 | 10.99 | 0.44 | 8.74 | 12.87 | 6 | 10.40 | 0.01 | 8.78 | 14.14 |
ZyL | 8 | 8.61 | 0.51 | 6.80 | 12.58 | 6 | 9.15 | 0.99 | 8.40 | 13.65 |
ZyR | 9 | 10.11 | 0.58 | 7.55 | 14.29 | 6 | 9.91 | 1.05 | 9.17 | 13.31 |
CoL | 9 | 19.10 | 3.51 | 16.72 | 25.83 | 6 | 16.20 | 2.40 | 14.50 | 21.42 |
CoR | 9 | 20.52 | 1.22 | 12.25 | 25.90 | 6 | 15.20 | 0.44 | 14.89 | 22.85 |
AlL | 7 | 9.38 | 1.68 | 7.44 | 13.64 | 5 | 7.69 | 0.01 | 6.10 | 14.48 |
AlR | 7 | 9.75 | 0.30 | 9.50 | 11.29 | 5 | 8.29 | 0.66 | 5.34 | 11.54 |
MnmL | 9 | 2.84 | 0.31 | 2.33 | 3.71 | 6 | 2.52 | 0.08 | 2.02 | 3.76 |
MnmR | 9 | 3.39 | 0.32 | 2.72 | 4.04 | 6 | 2.49 | 0.37 | 2.13 | 3.33 |
NmL | 8 | 6.08 | 1.13 | 3.73 | 7.18 | 5 | 4.43 | 0.20 | 3.11 | 5.48 |
NmR | 9 | 6.08 | 0.83 | 3.35 | 9.15 | 3 | 4.41 | 0.06 | 2.80 | 4.45 |
FmoL | 9 | 14.41 | 0.73 | 4.96 | 18.84 | 6 | 9.62 | 4.97 | 6.10 | 13.26 |
FmoR | 9 | 14.49 | 3.45 | 5.73 | 17.74 | 6 | 9.46 | 4.60 | 6.20 | 14.13 |
Table 7
Facial soft tissue thicknesses of the male and female individuals that were obtained using the 3D measurement method for the 55–64 years age group (measurements in mm)
Landmarks | Male (n = 7) | Female (n = 6) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
n | \(\:\stackrel{-}{\varvec{\upchi\:}}\) | σ | Min | Max | n | \(\:\stackrel{-}{\varvec{\upchi\:}}\) | σ | Min | Max | |
Sg | 7 | 5.72 | 1.23 | 4.34 | 7.77 | 6 | 6.66 | 0.91 | 6.17 | 8.28 |
G | 7 | 5.42 | 1.45 | 3.99 | 7.22 | 6 | 6.60 | 0.37 | 6.31 | 7.13 |
N | 7 | 6.50 | 1.15 | 5.82 | 10.05 | 6 | 7.38 | 1.20 | 6.57 | 9.34 |
Mn | 6 | 2.79 | 1.05 | 2.10 | 4.61 | 5 | 2.79 | 0.82 | 1.95 | 3.85 |
Rhi | 6 | 3.58 | 0.22 | 1.81 | 3.77 | 5 | 2.55 | 0.49 | 2.07 | 2.89 |
FeL | 6 | 6.58 | 0.41 | 3.96 | 6.94 | 5 | 5.39 | 0.18 | 5.26 | 8.45 |
FeR | 6 | 6.46 | 0.82 | 4.94 | 7.23 | 6 | 6.41 | 0.40 | 5.16 | 7.61 |
SoL | 7 | 10.22 | 0.84 | 7.14 | 12.17 | 5 | 9.32 | 0.93 | 8.66 | 11.19 |
SoR | 7 | 9.92 | 0.61 | 8.38 | 13.03 | 6 | 7.34 | 2.02 | 5.91 | 11.22 |
SkL | 7 | 10.02 | 1.33 | 8.09 | 12.75 | 5 | 10.01 | 0.47 | 9.68 | 11.76 |
SkR | 7 | 10.89 | 1.46 | 8.85 | 13.21 | 6 | 7.73 | 2.90 | 5.68 | 10.96 |
OrL | 7 | 6.75 | 0.70 | 4.05 | 7.47 | 5 | 7.43 | 1.72 | 6.21 | 9.28 |
OrR | 7 | 8.16 | 1.89 | 4.31 | 9.99 | 6 | 7.40 | 0.24 | 6.76 | 10.42 |
SmcL | 5 | 29.99 | 3.90 | 24.18 | 33.83 | 5 | 28.38 | 2.52 | 26.59 | 30.90 |
SmcR | 6 | 30.50 | 0.81 | 25.81 | 31.41 | 4 | 30.78 | 0.01 | 30.77 | 33.20 |
FtL | 7 | 6.09 | 0.83 | 4.88 | 7.95 | 5 | 7.72 | 1.59 | 5.34 | 9.29 |
FtR | 7 | 6.36 | 1.29 | 5.09 | 8.15 | 6 | 7.07 | 0.51 | 5.84 | 7.07 |
FmtL | 7 | 6.81 | 0.86 | 5.20 | 8.87 | 5 | 8.15 | 1.03 | 7.38 | 9.84 |
FmtR | 7 | 8.32 | 1.03 | 5.40 | 11.16 | 6 | 9.42 | 0.97 | 7.79 | 10.19 |
EcL | 7 | 12.26 | 1.34 | 5.48 | 13.45 | 5 | 10.63 | 2.39 | 6.30 | 15.59 |
EcR | 7 | 11.80 | 0.96 | 6.46 | 12.79 | 6 | 10.99 | 3.56 | 9.38 | 17.59 |
MzL | 7 | 11.01 | 1.56 | 6.49 | 12.26 | 5 | 9.87 | 1.42 | 9.87 | 12.69 |
MzR | 7 | 11.17 | 0.59 | 8.25 | 11.82 | 5 | 10.53 | 1.54 | 10.53 | 14.16 |
ZmL | 6 | 14.93 | 1.82 | 11.30 | 16.00 | 5 | 14.49 | 1.37 | 14.49 | 17.59 |
ZmR | 6 | 16.42 | 1.13 | 14.21 | 19.05 | 5 | 17.23 | 0.73 | 16.57 | 19.01 |
JuL | 7 | 10.31 | 1.47 | 6.48 | 11.78 | 6 | 12.75 | 0.70 | 11.09 | 12.89 |
JuR | 7 | 10.69 | 1.34 | 7.40 | 11.84 | 6 | 12.92 | 0.53 | 11.67 | 12.92 |
ZyL | 7 | 10.13 | 2.52 | 6.08 | 12.82 | 6 | 9.66 | 1.06 | 9.66 | 12.17 |
ZyR | 7 | 10.53 | 2.30 | 6.26 | 13.14 | 6 | 9.62 | 1.73 | 9.62 | 14.20 |
CoL | 7 | 21.49 | 3.01 | 18.07 | 23.76 | 6 | 19.57 | 3.04 | 15.94 | 23.56 |
CoR | 7 | 22.25 | 3.81 | 16.41 | 24.66 | 6 | 19.87 | 2.80 | 16.29 | 24.02 |
AlL | 6 | 8.77 | 1.05 | 7.26 | 9.61 | 5 | 9.08 | 3.17 | 7.50 | 14.76 |
AlR | 6 | 9.45 | 1.88 | 7.57 | 11.33 | 5 | 7.93 | 1.44 | 7.93 | 11.89 |
MnmL | 7 | 3.10 | 1.14 | 1.92 | 4.42 | 4 | 3.07 | 0.45 | 2.43 | 3.07 |
MnmR | 7 | 3.40 | 1.02 | 1.62 | 4.56 | 6 | 3.11 | 0.05 | 1.83 | 3.11 |
NmL | 7 | 5.75 | 0.70 | 3.78 | 6.35 | 4 | 5.81 | 0.31 | 3.57 | 5.81 |
NmR | 7 | 6.84 | 0.62 | 4.33 | 7.53 | 6 | 5.67 | 1.44 | 2.64 | 5.67 |
FmoL | 7 | 13.49 | 1.52 | 8.97 | 14.89 | 5 | 13.40 | 6.00 | 4.91 | 15.39 |
FmoR | 7 | 11.84 | 3.00 | 9.86 | 15.29 | 6 | 10.46 | 3.26 | 5.85 | 17.63 |
Table 8
Facial soft tissue thicknesses of the male and female individuals that were obtained using the 3D measurement method for the ≥ 65 years age group (measurements in mm)
Landmarks | Male (n = 21) | Female (n = 25) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
n | \(\:\stackrel{-}{\varvec{\upchi\:}}\) | σ | Min | Max | n | \(\:\stackrel{-}{\varvec{\upchi\:}}\) | σ | Min | Max | |
Sg | 21 | 6.10 | 1.20 | 3.11 | 9.46 | 24 | 5.33 | 1.04 | 4.39 | 8.04 |
G | 20 | 5.88 | 1.00 | 3.54 | 9.23 | 24 | 5.66 | 0.64 | 3.86 | 7.61 |
N | 21 | 7.36 | 1.53 | 4.66 | 11.02 | 25 | 5.16 | 0.74 | 4.27 | 9.69 |
Mn | 17 | 2.96 | 0.61 | 1.98 | 4.95 | 22 | 2.15 | 0.23 | 1.73 | 4.01 |
Rhi | 17 | 3.42 | 0.58 | 2.29 | 4.21 | 21 | 2.60 | 0.15 | 2.01 | 3.29 |
FeL | 18 | 6.46 | 0.93 | 4.37 | 8.50 | 23 | 4.78 | 1.03 | 3.78 | 7.72 |
FeR | 19 | 5.75 | 1.02 | 3.56 | 8.38 | 25 | 5.35 | 0.96 | 3.28 | 8.26 |
SoL | 19 | 9.06 | 1.57 | 6.13 | 14.11 | 24 | 8.78 | 2.07 | 5.88 | 12.04 |
SoR | 21 | 8.41 | 1.04 | 7.15 | 13.82 | 25 | 8.43 | 1.95 | 5.00 | 11.15 |
SkL | 21 | 8.91 | 1.56 | 4.84 | 15.53 | 24 | 8.25 | 1.92 | 5.60 | 12.28 |
SkR | 20 | 8.89 | 1.73 | 6.62 | 14.02 | 24 | 7.62 | 2.13 | 4.83 | 11.49 |
OrL | 21 | 6.37 | 2.56 | 2.93 | 12.58 | 23 | 5.27 | 0.68 | 3.91 | 9.88 |
OrR | 21 | 6.65 | 1.85 | 2.53 | 12.03 | 24 | 5.29 | 1.23 | 3.92 | 10.21 |
SmcL | 18 | 25.16 | 2.56 | 21.09 | 36.11 | 16 | 28.76 | 1.76 | 22.11 | 32.94 |
SmcR | 17 | 26.09 | 3.45 | 22.62 | 36.16 | 17 | 27.82 | 2.63 | 24.79 | 33.61 |
FtL | 21 | 6.25 | 1.09 | 3.96 | 9.27 | 24 | 5.66 | 2.08 | 3.23 | 8.73 |
FtR | 19 | 5.66 | 1.05 | 3.63 | 8.45 | 25 | 5.76 | 1.43 | 2.82 | 7.90 |
FmtL | 21 | 6.00 | 1.66 | 3.74 | 10.54 | 24 | 4.72 | 0.53 | 3.89 | 12.73 |
FmtR | 19 | 6.62 | 2.23 | 3.45 | 9.94 | 25 | 5.63 | 0.73 | 3.70 | 12.69 |
EcL | 20 | 11.24 | 1.61 | 3.67 | 19.03 | 24 | 10.10 | 3.56 | 2.53 | 15.11 |
EcR | 19 | 10.51 | 2.33 | 5.80 | 17.77 | 25 | 10.22 | 2.75 | 2.08 | 16.36 |
MzL | 20 | 7.99 | 1.27 | 5.68 | 12.60 | 23 | 9.38 | 1.46 | 7.50 | 12.60 |
MzR | 18 | 8.46 | 1.32 | 5.52 | 13.38 | 23 | 10.48 | 0.93 | 7.39 | 12.99 |
ZmL | 19 | 12.20 | 1.80 | 9.25 | 18.89 | 23 | 12.32 | 1.25 | 9.66 | 18.01 |
ZmR | 14 | 14.65 | 2.35 | 11.69 | 18.08 | 21 | 14.72 | 1.09 | 10.08 | 20.11 |
JuL | 21 | 8.24 | 1.78 | 5.01 | 13.04 | 24 | 10.05 | 2.44 | 7.67 | 14.22 |
JuR | 20 | 8.07 | 1.24 | 5.91 | 14.05 | 25 | 10.30 | 1.80 | 6.52 | 14.84 |
ZyL | 21 | 7.14 | 1.29 | 5.45 | 12.89 | 24 | 9.17 | 1.45 | 5.44 | 15.72 |
ZyR | 21 | 6.98 | 0.83 | 5.27 | 14.77 | 25 | 9.60 | 1.28 | 3.81 | 16.16 |
CoL | 21 | 17.30 | 2.50 | 11.72 | 28.43 | 24 | 15.96 | 1.54 | 9.61 | 22.21 |
CoR | 21 | 16.81 | 0.90 | 12.48 | 25.86 | 24 | 15.18 | 1.64 | 7.04 | 25.98 |
AlL | 19 | 10.47 | 2.26 | 5.33 | 13.37 | 21 | 9.53 | 3.26 | 5.36 | 14.00 |
AlR | 16 | 10.52 | 1.28 | 5.91 | 14.00 | 20 | 8.16 | 2.05 | 5.96 | 14.47 |
MnmL | 19 | 2.80 | 0.73 | 2.01 | 5.19 | 22 | 2.11 | 0.39 | 1.47 | 4.33 |
MnmR | 18 | 3.13 | 0.86 | 2.08 | 4.61 | 25 | 2.11 | 0.38 | 1.35 | 5.00 |
NmL | 19 | 5.07 | 1.05 | 3.75 | 9.85 | 14 | 4.66 | 0.73 | 2.06 | 5.67 |
NmR | 17 | 5.76 | 1.12 | 3.61 | 8.80 | 16 | 4.02 | 0.68 | 2.38 | 5.57 |
FmoL | 21 | 10.14 | 3.33 | 3.77 | 23.07 | 24 | 8.96 | 3.98 | 2.51 | 15.98 |
FmoR | 20 | 9.86 | 2.90 | 2.46 | 17.25 | 25 | 8.01 | 4.88 | 2.09 | 15.37 |
Impact of age and sex variables
The p-values of the age and sex variables and the two factor-between-subject effects are presented in Table 9. The mean depth values between the age groups varied considerably in the cheek and orbital regions for both males and females (Figs. 3 and 4). However, only the left orbitale, left mid-zygomatic and right condylion were found to be influenced by increasing age at the statistically significant level (p < 0.05). The mean FSTDs of females slightly increased with age. The largest depth values were generally reported from the 55–64 years age group. After this age group, the depth scores decreased again. On the contrary in females, there was not a specific age-related increasing or decreasing pattern in males.
Fig. 3
The line graph summarises the mean FSTDs of male subjects, including all five age groups
Fig. 4
The line graph summarises the mean FSTDs of female subjects, including all five age groups
The mean depth differences indicated that the FSTDs of males were slightly higher than those of females. Only the cheek region tends to be thicker in females, especially the submandibular curvature and mid-zygomatic. Despite the depth differences obtained between males and females, only at the left orbitale and right zygomaxillare was sexually dimorphic at the statistically significant level (p < 0.05). Similarly, the combined impact of age and sex variables was limited on the FSTDs of the Greek population sample (Table 9).
Table 9
The outcomes of the MANOVA test with the two factor-between-subject effects
Landmarks | Age | Sex | Age*Sex |
|---|---|---|---|
Sg | 0.18 | 0.26 | 0.07 |
G | 0.27 | 0.55 | 0.14 |
N | 0.57 | 0.98 | 0.30 |
Mn | 0.64 | 0.48 | 0.42 |
Rhi | 0.30 | 0.65 | 0.62 |
FeL | 0.78 | 0.45 | 0.86 |
FeR | 0.87 | 0.64 | 0.79 |
SoL | 0.87 | 0.63 | 0.87 |
SoR | 0.97 | 0.78 | 0.88 |
SkL | 0.90 | 0.67 | 0.78 |
SkR | 0.69 | 0.68 | 0.66 |
OrL | 0.01** | 0.03* | 0.00** |
OrR | 0.16 | 0.27 | 0.04* |
SmcL | 0.12 | 0.37 | 0.14 |
SmcR | 0.08 | 0.32 | 0.05* |
FtL | 0.36 | 0.46 | 0.10 |
FtR | 0.56 | 0.66 | 0.29 |
FmtL | 0.90 | 0.63 | 0.78 |
FmtR | 0.86 | 0.58 | 0.53 |
EcL | 0.93 | 0.63 | 0.94 |
EcR | 0.85 | 0.99 | 0.39 |
MzL | 0.04* | 0.52 | 0.24 |
MzR | 0.18 | 0.75 | 0.20 |
ZmL | 0.33 | 0.52 | 0.21 |
ZmR | 0.10 | 0.04* | 0.41 |
JuL | 0.41 | 0.73 | 0.84 |
JuR | 0.21 | 0.83 | 0.47 |
ZyL | 0.17 | 0.52 | 0.64 |
ZyR | 0.13 | 0.38 | 0.35 |
CoL | 0.30 | 0.15 | 0.26 |
CoR | 0.02* | 0.01 | 0.14 |
AlL | 0.94 | 0.70 | 0.65 |
AlR | 0.43 | 0.09 | 0.35 |
MnmL | 0.37 | 0.41 | 0.59 |
MnmR | 0.14 | 0.08 | 0.36 |
NmL | 0.60 | 0.52 | 0.91 |
NmR | 0.52 | 0.08 | 0.25 |
FmoL | 0.95 | 0.64 | 0.66 |
FmoR | 0.95 | 0.76 | 0.37 |
Bilateral asymmetry
The comparison between the landmark pairs is summarised in Table 10. The FSTDs measured from the right side of the face were slightly larger than the left side in males. The greatest difference was measured between the left and right zygomaxillare (1.89 mm). Only the bilateral differences between the frontomalare temporale, zygomaxillare, mid-nasomaxillare, and nasomaxillare were statistically significant for males (p < 0.05).
In females, the left side of the face was larger than the right side. While the maximum depth difference was reported between the left and right zygomaxillare (1.14 mm), the minimum difference was recorded between the left and right frontal eminence, ectoconchion, and zygion (0.10 mm) (Table 10). The number of cranial landmarks that showed statistically significant bilateral asymmetry was higher in females than in males. The absolute mean depth differences in both males and females suggested that the bilateral asymmetry calculated between the landmark pairs was negligible.
Table 10
Paired sample t-test and Wilcoxon signed-rank test, comparing the FSTD values between the left and right sides of the face
Landmarks | Male | Female | ||
|---|---|---|---|---|
Diff. | Sig. | Diff. | Sig. | |
FeL/FeR | 0.07 | 0.20 | 0.10 | 0.62 |
SoL/SoR | 0.04 | 0.68 | 0.37 | 0.32 |
SkL/SkR | 0.08 | 0.30 | 0.42 | 0.27 |
OrL/OrR | 0.22 | 0.10 | 0.18 | 0.99 |
SmcL/SmcR | 1.21 | 0.07 | 0.57 | 0.45 |
FtL/FtR | 0.13 | 0.22 | 0.13 | 0.25 |
FmtL/FmtR | 0.57 | 0.02* | 0.64 | 0.00** |
EcL/EcR | 0.06 | 0.55 | 0.10 | 0.29 |
MzL/MzR | 0.55 | 0.43 | 0.43 | 0.03* |
ZmL/ZmR | 1.89 | 0.00** | 1.14 | 0.00** |
JuL/JuR | 0.23 | 0.21 | 0.19 | 0.01** |
ZyL/ZyR | 0.39 | 0.06 | 0.10 | 0.02* |
CoL/CoR | 0.10 | 0.85 | 0.29 | 0.05* |
AlL/AlR | 0.20 | 0.44 | 0.32 | 0.85 |
MnmL/MnmR | 0.10 | 0.05* | 0.23 | 0.85 |
NmL/NmR | 0.09 | 0.00** | 0.21 | 0.39 |
FmoL/FmoR | 0.13 | 0.60 | 0.14 | 0.12 |
Interpopulation variation
The FSTDs of the Greek population were compared with the reported mean FSTDs of five populations (Tables 11, 12, 13, 14 and 15). The mean facial depth values of Greek males were slightly larger than those of the Cretan males, ranging between 0.06 and 2.04 mm. The highest depth difference was reported from the nasomaxillare. The FSTD differences between the supraglabella, mid-nasal, rhinion, supraorbital, orbitale, frontomalare temporale, ectoconchion, mid-zygomatic, alare, mid-nasomaxillare and nasomaxillare were at a statistically significant level (p < 0.05). The mean FSTD scores of the Cretan females were slightly lower than those of the Greek females at most of the landmarks (Table 11). The largest mean depth difference was calculated between the ectoconchion (1.86 mm). The smallest difference, however, was reported from the rhinion (0.02 mm). Apart from the nasion, rhinion, zygomaxillare and frontomalare orbitale, the p-values of all the landmarks showed statistically significant differences (p < 0.05).
The FSTD of the Greek males were negligible larger than those of the Turkish males. The highest depth difference was recorded at the supraorbital with 2.81 mm (Table 12). In contrast, the smallest difference was reported from the glabella with 0.12 mm. Apart from the glabella, all the facial depth values of the Greek males were found to be statistically different from those of the Turkish males. Similarly, the mean FSTD values of Greek females were slightly larger than those of the Turkish females at all the landmarks, apart from the nasion. The mean FSTD values of the nasion were 0.40 mm larger in the Turkish females. The mean value of the glabella was measured the same in both populations, which was 6.03 mm. This comparison showed that the mean FSTDs of Greek females were statistically different from those of Turkish females at the eight landmarks (p < 0.05). Only the p-value of the glabella was above 0.05 (p > 0.05).
Similar results were obtained from the comparison made between the Greek and Korean male subjects. The mean FSTD values of Greek males were larger than those of the Korean males. The results of one sample t-test showed that all the landmarks showed statistically significant differences at the significance level of p < 0.01, except the frontal eminence (Table 13). The mean differences between Greek and Korean females were compatible with those of males. The mean depth values of the Korean females were generally smaller than those of the Greek females. The largest variance was reported from the supraorbital (2.00 mm), which was a still negligible difference. All the p-values of landmarks, apart from the frontal eminence and mid-zygomatic, were below 0.05.
The comparison between the Czech and Greek males indicated that the FSTD values were higher in the Czech adult males than the Greek males at most of the landmarks. The mean depth differences of males varied between 0.01 mm and 4.82 mm. The highest depth differences were measured at the left and right ectoconchion (4.82 mm and 4.77 mm respectively) (Table 14). The p-values of nasion, left and right orbitale, left and right ectoconchion, right jugale, and left and right alare were lower than 0.05, indicating that the mean difference at these landmarks was statistically significant. On the contrary of the mean male depth values, the mean FSTD scores were larger in the Greek females than the Czech females. The largest depth differences were measured at the left and right ectoconchion (3.81 mm and 3.69 mm respectively) in females as well. Apart from the glabella, rhinion, left zygion, and left and right jugale, all the landmarks showed statistically significant differences (p < 0.01 and 0.05).
The FSTD datasets of male and female subjects were not reported separately for the French population sample [17]. Therefore, the comparison across the Greek and French subjects was made using the average depth values of all subjects. The French subjects showed the larger FSTD scores at most landmarks (Table 15). However, the mean depth differences were not significant and did not exceed 2.58 mm. Of 16 landmarks, the p-values of 13 landmarks were statistically significant, with a significance level of p < 0.01 and p < 0.05.
Table 11
Comparison of the mean FSTD values of the present study with the Cretan [37] population for male and female subjects
Landmarks | Male | Female | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
n | Greek | n | Cretan | Sig | n | Greek | n | Cretan | Sig | |
Sg | 50 | 6.54 | 32 | 5.7 | 0.00** | 48 | 5.93 | 32 | 4.8 | 0.00** |
G | 49 | 6.28 | 32 | 5.9 | 0.06 | 49 | 6.03 | 32 | 5.5 | 0.00** |
N | 50 | 8.02 | 32 | 7.8 | 0.35 | 50 | 6.76 | 32 | 6.7 | 0.76 |
Mn | 41 | 3.30 | 30 | 4.5 | 0.00** | 44 | 2.57 | 29 | 3.0 | 0.00** |
Rhi | 42 | 3.30 | 30 | 3.8 | 0.00** | 44 | 2.62 | 24 | 2.6 | 0.70 |
Fe | 45 | 6.27 | 32 | 6.1 | 0.42 | 49 | 5.55 | 32 | 3.9 | 0.00** |
So | 49 | 9.75 | 32 | 8.6 | 0.00** | 50 | 8.40 | 32 | 7.2 | 0.00** |
Sc | 47 | 10.04 | 32 | 10.4 | 0.14 | 49 | 8.52 | 32 | 9.3 | 0.00** |
Or | 50 | 7.41 | 30 | 9.1 | 0.00** | 49 | 6.52 | 24 | 7.4 | 0.00** |
Smc | 39 | 29.12 | 20 | 29.9 | 0.19 | 36 | 28.54 | 10 | 27.3 | 0.01** |
Ft | 48 | 6.19 | 32 | 6.4 | 0.32 | 50 | 5.77 | 32 | 4.6 | 0.00** |
Fmt | 48 | 7.91 | 32 | 7.0 | 0.00** | 50 | 7.61 | 32 | 6.0 | 0.00** |
Ec | 45 | 10.61 | 32 | 9.4 | 0.03* | 50 | 9.56 | 29 | 7.7 | 0.00** |
Mz | 41 | 9.49 | 24 | 11.2 | 0.00** | 46 | 10.39 | 19 | 11.4 | 0.00** |
Zm | 36 | 15.75 | 20 | 16.6 | 0.28 | 44 | 15.38 | 13 | 14.8 | 0.14 |
Ju | 48 | 9.66 | 32 | 9.6 | 0.84 | 50 | 10.62 | 31 | 9.9 | 0.00** |
Zy | 49 | 9.11 | 32 | 9.7 | 0.09 | 50 | 9.89 | 31 | 9.0 | 0.01** |
Al | 39 | 9.75 | 21 | 10.9 | 0.00** | 42 | 9.23 | 10 | 10.0 | 0.02* |
Mnm | 45 | 3.24 | 30 | 2.9 | 0.00** | 50 | 3.00 | 27 | 2.1 | 0.00** |
Nm | 44 | 6.04 | 29 | 4 | 0.00** | 33 | 4.20 | 20 | 2.8 | 0.00** |
Fmo | 49 | 11.09 | 32 | 10.2 | 0.10 | 50 | 10.04 | 32 | 10.1 | 0.91 |
Table 12
Comparison of the mean FSTD values of the present study with the Turkish [22] population for male and female subjects
Landmarks | Male | Female | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
n | Greek | n | Turkish | Sig. | n | Greek | n | Turkish | Sig. | |
Sg | 50 | 6.54 | 42 | 4.21 | 0.00** | 48 | 5.93 | 42 | 3.94 | 0.00** |
G | 49 | 6.28 | 42 | 6.40 | 0.53 | 49 | 6.03 | 42 | 6.03 | 0.99 |
N | 50 | 8.02 | 42 | 7.32 | 0.00** | 50 | 6.76 | 42 | 7.16 | 0.03* |
Rhi | 42 | 3.30 | 42 | 2.97 | 0.00** | 44 | 2.62 | 42 | 2.38 | 0.00** |
Fe | 45 | 6.27 | 42 | 4.48 | 0.00** | 49 | 5.55 | 42 | 4.02 | 0.00** |
So | 49 | 9.75 | 42 | 6.94 | 0.00** | 50 | 8.40 | 42 | 6.17 | 0.00** |
Or | 50 | 7.41 | 42 | 5.92 | 0.00** | 49 | 6.52 | 42 | 5.47 | 0.00** |
Mz | 41 | 9.49 | 42 | 7.90 | 0.00** | 46 | 10.39 | 42 | 9.40 | 0.00** |
Zy | 49 | 9.11 | 42 | 7.83 | 0.00** | 50 | 9.89 | 42 | 8.06 | 0.00** |
Table 13
Comparison of the mean FSTD values of the present study with the Korean [23] population for male and female subjects
Landmarks | Male | Female | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
n | Greek | n | Korean | Sig. | n | Greek | n | Korean | Sig. | |
Sg | 50 | 6.54 | 50 | 5.3 | 0.00** | 48 | 5.93 | 50 | 4.8 | 0.00** |
G | 49 | 6.28 | 50 | 5.6 | 0.00** | 49 | 6.03 | 50 | 5.3 | 0.00** |
N | 50 | 8.02 | 50 | 6.4 | 0.00** | 50 | 6.76 | 50 | 5.4 | 0.00** |
Rhi | 42 | 3.30 | 50 | 2.3 | 0.00** | 44 | 2.62 | 50 | 2.2 | 0.00** |
Fe | 45 | 6.27 | 50 | 6.2 | 0.70 | 49 | 5.55 | 50 | 5.4 | 0.39 |
So | 49 | 9.75 | 50 | 7.2 | 0.00** | 50 | 8.40 | 50 | 6.4 | 0.00** |
Mz | 41 | 9.49 | 50 | 8.6 | 0.00** | 46 | 10.39 | 50 | 10.2 | 0.41 |
Zy | 49 | 9.11 | 50 | 8.1 | 0.00** | 50 | 9.89 | 50 | 8.7 | 0.00** |
Table 14
Comparison of the mean FSTD values of the present study with the Czech [19] population for male and female subjects
Landmarks | Male | Female | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
n | Greek | n | Czech | Sig. | n | Greek | n | Czech | Sig. | |
G | 49 | 6.28 | 56 | 6.29 | 0.94 | 49 | 6.03 | 46 | 6.01 | 0.80 |
N | 50 | 8.02 | 56 | 9.41 | 0.00** | 50 | 6.76 | 46 | 8.25 | 0.00** |
Rhi | 42 | 3.30 | 56 | 3.11 | 0.10 | 44 | 2.62 | 46 | 2.61 | 0.81 |
OrL | 50 | 7.18 | 56 | 8.37 | 0.00** | 47 | 6.34 | 46 | 7.47 | 0.00** |
OrR | 50 | 7.41 | 56 | 8.35 | 0.01** | 49 | 6.52 | 46 | 7.22 | 0.01** |
EcL | 49 | 10.67 | 56 | 5.85 | 0.00** | 48 | 9.66 | 46 | 5.85 | 0.00** |
EcR | 45 | 10.61 | 56 | 5.84 | 0.00** | 50 | 9.56 | 46 | 5.87 | 0.00** |
ZyL | 49 | 8.71 | 56 | 8.77 | 0.85 | 49 | 9.79 | 46 | 9.19 | 0.09 |
ZyR | 49 | 9.11 | 56 | 8.49 | 0.08 | 50 | 9.89 | 46 | 9.17 | 0.05* |
JuL | 49 | 9.42 | 56 | 9.72 | 0.30 | 49 | 10.43 | 46 | 10.27 | 0.56 |
JuR | 48 | 9.65 | 56 | 10.36 | 0.00** | 50 | 10.62 | 46 | 10.56 | 0.82 |
AlL | 42 | 9.95 | 56 | 12.35 | 0.00** | 43 | 9.55 | 46 | 10.69 | 0.00** |
AlR | 39 | 9.75 | 56 | 12.02 | 0.00** | 42 | 9.23 | 46 | 10.62 | 0.00** |
Table 15
Comparison of the mean FSTD values of the present study with the French [17] population
Landmarks | n | Greek | n | French | Sig. |
|---|---|---|---|---|---|
G | 98 | 6.16 | 366 | 6.5 | 0.00** |
N | 100 | 7.39 | 469 | 8.2 | 0.00** |
Mn | 85 | 2.92 | 321 | 5.5 | 0.00** |
Rhi | 86 | 2.96 | 459 | 3.0 | 0.54 |
Sc | 96 | 9.26 | 373 | 9.7 | 0.03* |
Or | 99 | 6.97 | 371 | 8.6 | 0.00** |
Smc | 75 | 28.84 | 330 | 28.2 | 0.08 |
Ft | 98 | 5.98 | 361 | 6.7 | 0.00** |
Fmt | 98 | 7.76 | 242 | 9.1 | 0.00** |
Ec | 95 | 10.06 | 372 | 7.8 | 0.00** |
Zm | 80 | 15.55 | 353 | 15.1 | 0.09 |
Ju | 98 | 10.15 | 245 | 10.9 | 0.00** |
Zy | 99 | 9.50 | 364 | 10.0 | 0.04* |
Co | 99 | 17.73 | 215 | 16.7 | 0.02* |
Mnm | 95 | 3.11 | 328 | 3.5 | 0.00** |
Nm | 77 | 5.25 | 457 | 4.9 | 0.04* |
Discussion
Accuracy of head and skull segmentations
Following the Daubert v. Merrell Dow Pharmaceuticals, Inc. [58] case, new standards for scientific evidence were introduced in forensic disciplines. To be considered scientifically valid, evidence must be testable, reproducible with a known error rate, subject to peer review, and accepted by the respective scientific community [39]. These admissibility criteria led many forensic disciplines to question the validity of their existing methods. Forensic anthropology was one such field that sought to establish standards. However, there are still aspects of the discipline, such as estimating age and approximating faces, that rely on subjective interpretation. To address this issue and improve the reliability of the methods, Christensen and Crowder [40] recommended reporting the repeatability of the methods used.
The study found that the 3D head and skull models were accurately reproduced using the Histogram method. This assessment was crucial in ensuring the precision of the re-segmentations prior to the depth measurements. The results of this study have significant implications, as the Histogram method can be applied to other fields that require the segmentation of CT images. In FSTD studies, the half maximum height (HMH) protocol is typically used for segmentation. However, this method is time-consuming as it requires a lengthy calculation process of HMH values for each individual, as well as an additional edge correction. On the other hand, the Histogram method also requires edge correction when necessary but is more time-efficient as it does not require quantifying HMH values.
Intra- and inter-observer error
Some degree of measurement error is always expected from the studies in which anthropometric measurements are involved [32]. In the FSTD literature, the technical error of the measurements is rarely reported. Stephan et al. [41] suggested calculating TEM and rTEM values to increase the integrity between the FSTD studies. In this study, the facial soft tissue depth measurements were tested for repeatability and reliability within and between observers. The results of the TEM, rTEM and R scores of the intra-observer error were within the acceptable level of accuracy.
All the measurements carried out were subject to skull positioning-associated errors. Since the measurements were taken from the 3 different planes (Fig. 5), the skulls were rotated, and their positions were re-adapted each time based on the location of landmarks. This methodological requirement might have led to a divergence between the repeating measurements.
Fig. 5
The measurement directions of the landmarks. Mid-sagittal (A) and a number of bilateral (B) landmarks were measured when the skull was in the sagittal plane. The rest of the landmarks were measured when the skull was in the coronal (C) and/or sagitto-coronal (D) planes
Bookstein [42] divides landmarks into three types based on their anatomical shape and position. Type I landmarks are described as discrete juxtapositions of tissues. These landmarks are found where the three structures and/or sutures meet, such as the nasion, nasomaxillofrontale, nasomaxillare and zygomaxillare. Type II landmarks are defined as maxima of curvature or local morphometric processes, involving the tips of the processes, and valleys of invaginations such as the jugale, supraglabella, rhinion and ectoconchion [43]. Type III landmarks are, however, defined as those found at the extremal locations, such as the glabella, frontomalare orbitale, frontotemporale, zygion, and sub-maxillar curvature. Studies showed that Type I landmarks are identified with high accuracy, whereas Type II and III landmarks are prone to misidentification [38, 43‐45]. Despite these findings, the rTEMs of the zygomaxillare, mid-nasomaxillare and nasomaxillare (Type I landmarks) of this study appeared to be high. This result could be explained by the failure of reconstruction of the zygomaticomaxillary and nasomaxillary sutures. This issue might have caused inconsistent identification of the respective landmarks on these sutures. Bertoglio et al. [47] stated that the non-metric traits and facial sutures, which are formed by the thin layer of compact bone, are less likely to be visible on 3D models (especially maxillary bone). Additionally, Kim et al. [46] found that the quality of images decreases with increasing slice thickness, causing less precise 3D models. Accordingly, the images with the 0.5 mm (and below) slice thickness provide the best image quality [46]. Although the CT slice thickness of this study (2.4 mm) was under 5 mm, it could be still considered thick to recreate the facial sutures accurately when their size is taken into account.
Influence of age and sex on FSTD
The impact of age on FSTDs has been studied by several studies, which have reported varying results. Wilkinson [14] revisited previously published FSTD studies and concluded that the influence of age on the FSTD was variable. As the general tendency suggests, the depth at the mouth and lower cheek region decreases, whereas the depth at the chin and brow region increases with age [14, 59]. Both increasing and decreasing trends are expected from the facial depth since the volume of the superficial and deep facial fat compartments change with age [52, 53].
In this study, the greatest depth values of males were mainly reported from the 35–44 years age group. Yet, after this age group, the mean values did not display any regular pattern, they fluctuated. Therefore, it was not possible to relate the mean facial depth of any landmarks and/or facial regions (e.g., cheek, orbital, nasal) with increasing age. This fluctuation might be attributed to the varying body weight of the individuals measured for this study. The impact of body weight on FSTD has been analysed by several researchers [26, 49, 51]. Although there is a slight deviation across the results of the studies, the general agreement is that FSTD increases as body weight increases. Since the individuals in this study could not be categorised based on their BMIs, the age groups of both sex categories might have consisted of subjects with mixed body types (e.g., slender, normal and overweight). This issue might have obscured the real impact of age and sex variables on the FSTD. In females, the FSTD values increased with age at the nasal, orbital and cheek regions. The depth values mostly peaked at the 55–64 years age group, then decreased.
The researchers have investigated the influence of sex on FSTD and found that males have larger facial depth than females [20, 23, 26, 27]. In particular, the brow, jaw and mouth regions are thicker in males, whereas the cheek region is thicker in females [14, 26, 27]. The findings of the current study also demonstrated that male depth values were generally higher than those of female depth values.
Cha [54] found some depth differences statistically significant between sexes, yet not exceeding 2 mm. This difference was attributed to the facial fat compartments, as well as the age-dependent decrease in skin thickness. In general, the skin thickness appears to be larger in males than in females. With increasing age, the skin undergoes thinning, dryness and rough texture. In addition, the density of collagen in the skin, which is the main contributor to skin thickness, decreases with advancing age [55]. This decrease occurs at different rates in males and females [56]. While male skin demonstrates progressive thinning with ageing, female skin remains steady until around middle age, and then rapid alteration occurs [56, 57]. In this study, the mean FSTDs of males showed inconsistently increasing and decreasing trends between the age groups. Therefore, there was not enough evidence to relate the depth changes to a possible impact of skin thinning in males. The female depth scores, on the other hand, decreased after the 55–64 years age group. The decrease in mean FSTDs of females might be explained by skin thinning along with the decrease in the volume of the facial fat masses and attenuation of the facial soft tissues.
Although the depth differences between sexes were reported as statistically significant by some scholars, the depth variation was reported as insignificant [17, 25, 28]. As a result, they recommend combining depth measurements to create a general population database, as sex has little impact on facial soft tissue depths. This study’s findings are in agreement with these previous studies [17, 28], showing that the differences in depth between landmarks were mostly insignificant. Furthermore, the variation in depth within and across sex categories was low. Therefore, this study supports the idea put forward by Stephan and Simpson [25] that a general dataset can be generated for both males and females since sex has no significant impact on facial soft tissue depth in this population sample.
Bilateral asymmetry
Contrary to the general belief, the human body rarely displays complete symmetry. The degree of asymmetry may vary, and it may not always be evident to the naked eye [48]. The assessment of facial asymmetry showed variety across the FSTD studies, yet this factor was not taken into account in all studies. The general agreement in the FSTD literature regarding facial asymmetry is that statistically significant variation between the left and right sides of the face might be observed at varying numbers of landmarks, yet the absolute mean differences between the bilateral landmarks are insignificant [23, 25, 49]. Therefore, researchers either averaged bilateral depth values and reported one single value for each bilateral landmark [25, 50], or measurements were taken only from the left side of the face by anthropological convention [16, 21]. Alternatively, although the landmarks at both sides were measured, only the right side of the face was reported since the absolute mean depth differences were insignificant [18, 37, 49].
In this study, the right side of the face was found to be slightly thicker in males than the left side. In females, however, the left side was insignificantly larger than the right side. This bilateral variation appeared to be small. Thus, measuring one side of the face and using the FSTD dataset of only one side would not compromise facial approximations.
Interpopulation variation
Researchers reported varied inter-population depth results over time. Some studies suggested establishing population-specific datasets since significant depth variations were acquired across the studies [21, 26, 37], whereas other studies claimed that there are negligible differences between the datasets. These differences have little practical significance [16, 17, 20, 28].
An extensive FSTD comparison was made by Stephan and Simpson [28] by using the datasets of 55 populations. As they reported, the FSTD scores varied across the populations. However, they also reported that variations within the populations were greater than the variations across the populations. The difference obtained cannot be attributed only to the inter-population difference since all the studies investigated used different data-obtaining methods (e.g., ultrasound, radiograph, MRI, CT, CBCT), measurement methods (i.e., measurements from cadavers, measurements from 2D images and measurements from 3D models), and the FSTDs were measured by different examiners. Stephan and Simpson [28] concluded that inter-population FSTD differences are of minor practical importance for facial approximation. Therefore, all the population datasets should be pooled in order to increase the sample size and establish one statistically powerful dataset for all populations.
Although there were statistically significant differences in the facial depth at most landmarks, the variation in depth across populations was minimal for both males and females in this study. This result was in accordance with the results of other studies [16, 17, 24, 28]. The mean depth differences obtained from geographically close and distant populations to the Greek population did not vary considerably. This suggests that there is no positive correlation between geographical distance and facial depth. Although the population comparison of this study was made between the studies that used similar data-obtaining methods and measurement protocols, the impact of the examiner and individual facial depth variation should be taken into account. As suggested by Stephan and Simpson [28], pooling the FSTD datasets of all other populations might generate a single, comprehensive, and statistically powerful dataset.
Conclusion
The facial soft tissue dataset is a significant guideline for facial approximation and superimposition methods [25, 28]. Although the results of a number of other studies [17, 20, 24, 28] reported small inter-population variations, researchers are still encouraged to establish a population-specific dataset so that a pooled dataset could be built [28]. The result of this study contributes to the craniofacial approximation literature by compiling the FSTD dataset for the Greek population for the first time.
The database generated in this study is significant not only for the identification of unknown Greek individuals but also for a broader perspective, it is important to contribute literature to create a more comprehensive dataset that can be used for all populations. The identification of the victims of migration, natural disasters, crime and armed conflicts is a significant problem due to the lack of antemortem data. In that case, performing forensic anthropological methods and facial approximation of victims might assist in identifying the remains. When population affinity cannot be accurately set, a weighted FSTD database could be used.
Acknowledgements
The author (GC) would like to thank Dr Linda Fibiger and Dr Jonny Geber for their invaluable support and supervision during this research.
Declarations
Ethical approval
This study was performed in line with the principles of the Declaration of Helsinki. The ethical committees of Athens Medical Centre granted the acquisition of the anonymised CT scans (23/03/2020) and the University of Edinburgh (School of History, Classics and Archaeology) gave their approval for conducting this study (19/02/2020).
Informed consent
Not applicable.
Conflict of interest
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
