7.1 Introduction
7.2 Study on the Performance of Different Craniofacial Superimposition Approaches
7.2.1 Experimental Study
Case | # Skulls | # Photographs | # Candidates | # SFO’S | # Positive cases | |
---|---|---|---|---|---|---|
Male set | Female set | |||||
N1 | 1 | 3 | 3 | 3 | 1 | 0 |
N2 | 1 | 3 | 3 | 3 | 1 | 1 |
N3 | 1 | 3 | 3 | 3 | 0 | 1 |
N4 | 1 | 3 | 3 | 3 | 0 | 0 |
N5 | 4 | 4 | 4 | 16 | 2 | 2 |
N6a
| 1 | 2 | 1 | 2 | 0 | 0 |
N7a
| 1 | 2 | 1 | 2 | 2(1)a
| 2(1)a
|
TOTAL | 7 | 20 | 18 | 32(30)a
| 6(5)a
| 6(5)a
|
-
No previous experience and no CFS-related training.
-
No previous experience but CFS-related training.
-
Short previous research experience and CFS-related training.
-
Moderate previous experience with CFS real cases and CFS-related training.
-
Broad experience with CFS real cases.
Participant number | P1 | P2 | P3 | P4 | P5 | P6 | P7 | P8 | P9 | P10 | P11 | P12 | P13 | P14 | P15 | P16 | P17 | P18 | P19 | P20 | P21 | P22 | P23 | P24 | P25 | P26 | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
CFS experience | 4 | 3 | 3 | 3 | 5 | 4 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 3 | 2 | 4 | 4 | 1 | 4 | 1 | 4 | 5 | 3 | 5 | 5 | 5 | |
Submitted results | Males | ✓ | ✓ | ✗ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
Females | ✓ | ✓ | ✓ | ✗ | ✓ | ✗ | ✗ | ✗ | ✗ | ✗ | ✗ | ✗ | ✗ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✗ | ✓ | ✗ | ✗ | ✓ | ✓ |
Participant number and global performance | Software and equipment | Skull-face overlay | Decision making |
---|---|---|---|
P2 Both datasets 91.43% | Face2Skull™ | Automatic SFO by matching cranial and facial landmarks (manually located). Manual refinement of the overlay when needed. | |
P19 Both datasets 83.01% | 3D-TADD | Used 3D-TADDsoftware to scale the photograph and position the 3D model. | Morphological and landmark assessment. Anatomical points and contours of the head, the middle and lower part of the face, eyebrows, ears lobes, the line between the lips, the wings of the nose and supramental furrow are automatically detected. Manual refinement of points and contours when needed. |
Participant number and global performance | Software and equipment | Skull-face overlay | Anatomical consistency assessment criteria | Observation |
---|---|---|---|---|
P4 Male dataset 94.29% | Freeform Modelling Plus Phantom® Desktop™ Haptic Device Adobe Photoshop™ Wacom tablet | The cranium was orientated first in a trial and error manner, by means of moving and rotating the stylus of the haptic device. Starting points for the alignment were eyes/orbits in conjunction with either the maxillary teeth, if visible, or the nasal bridge. Follow “fuzzy landmarks” and morphology for orientation. The mandible was orientated in compliance with the facial expression and the limits of a physiologically natural position. | The resulting images were then imported into Adobe Photoshop and outline tracings of the skull and face were produced using Wacom tablet. | The software has two variables to “play” with the camera projection. |
P3 Female dataset 91.43% | Freeform Modelling Plus Phantom® Desktop™ Haptic Device Adobe Photoshop™ | Manual orientation with software and haptic device. Anatomical criteria of morphological assessment taken into account during this process. | Morphological assessment based on (Wilkinson 2006). | Haptic device used for moving the mandible. |
P18 Both datasets 87.69% | Blender Adobe Photoshop™ | Manual orientation using Blender and Photoshop. | Morphological assessment. | |
P14 Both datasets 85.71% | Geomagic Studio10 (3D software) Adobe Photoshop™ CS6 | Geomagic Studio used to articulate cranium and mandible. Orientation of the skull by trial and error manipulation of the 3D skull model using Photoshop | Landmark and morphological consistency analysis. | Consider first biological profile, asymmetries, and pathologies. Give special consideration to teeth comparison. |
P20 Male dataset 81.66% | Autodesk 3D Studio MAX Photoshop™ CS6 | Orientation (yaw, pitch and roll) setting in the same plane porion, tragus, Whitnall’s tubercle, and ectocanthion (Fenton et al. 2008). | Morphological (outlines) and landmark matching assessment. | Asymmetries are key signs for individualization. |
P21 Male dataset 81.82% | Adobe Photoshop™ | Both 3D model and AM photo oriented and resized in Photoshop. Separate use of cranium and mandible. | Morphological and landmark assessment. | |
P5 Both datasets 80.88% | Adobe Photoshop ™ CS3. 3D Max Studio 2011 | Recommendations by Lan and Cai (1993). Morphological assessment. | ||
P1 Both datasets 79.10% | RapidForm 2006 Adobe Photoshop™ | 3D skull model oriented by simultaneous view of the model and the AM photo. Resize of the image according to cranial and facial landmarks. | Followed criteria described in Navarro-Merino (2011). | |
P17 Both datasets 77.19% | Craniofacial correspondence analysis program (in-house software) | The program represents the algorithm of transformations from dimensional and descriptive traits of a skull into the corresponding parameters of the face. | The program of craniofacial correspondence based on the data on soft facial tissue thickness and on the analysis of relationships between facial and skull features, correlative and regression analysis of the facial and skull features, as well as a descriptive analysis of indexes and variations of categorical and quantitative characteristics. | |
P15 Both datasets 76.81% | Geomagic Studio software 3D Rugle | The 3Dscan is rotated manually (using 3D Rugle) through all possible axes (X,YandZ) to Frankfort horizontal plane. The orientation of the skull is then electronically realigned so that the anatomical position becomes the starting position. | Morphological and landmark assessment. | Geomagic is also used to complete the mandible articulation. Each photograph is cropped and resized to maximize the dimensions of the face in the frame. |
P22 Both datasets 71.43% | Claytools Geomagic, Inc. 3D Rugle Adobe Photoshop™ Elements | The obj 3D data both of “cranium” and “mandible” were imported to a 3D modeling tool, Claytools (Geomagic, Inc.). A craniofacial superimposition was performed with “3D Rugle” software. These are imported to an Adobe Photoshop Elements software and superimposed again. | Morphological and landmark assessment (measure distances between landmarks). | A perspective effect can be virtually applied to the 3D skull data using this software (3D Rugle). |
P16 Both datasets 67.24% | Rhinoceros© or Autodesk 3D Studio© software Adobe Photoshop™ software | Right size is established from inter-Whitnall’s malar tubercular distance and/or glabella–gnathion distance. Orientation is established matching Whitnall’s malar tubercles–porion, where possible. Superimposition based on shapes. Landmarks help guiding the process | Morphological and landmark assessment. | Generally do not use 3D laser scan, because participants have access to a CT-scan with 0.5 mm isotropic voxel that helps in obtaining a very anatomically detailed 3D image. |
P23 Male dataset 55.88% | Geomagic Studio software 3D Studio Max software VAM software | Landmark-based orientation following Gordon and Steyn (2012). | Landmark-based consistency analysis following Gordon and Steyn (2012). | 0, 1, or 2 disagreements of landmarks are considered a positive match. 3 or more a considered a negative match. |
Participant number and global performance | Software and equipment | Skull-face overlay | Decision making |
---|---|---|---|
P24 Mala dataset 80.00% | Digital video mixer, digital video camera, and screen monitor | The orientation was done by aligning: Whitnall’s tubercle with ectocanthions; external auditory meatus with the tragus; vertical midline of the face using the glabella, nasion, anterior nasal spine, prosthion, menton, and pogonion. The size was adjusted taking soft tissue depth into consideration. | First, the correlation of the shape characteristics between facial skeleton and face in the photograph was compared. Then, morphological assessment was performed. |
P25 Both datasets 83.33% | Digital video mixer, digital video camera, and screen monitor AVerMedia TV Series Product | Skull: (1) orientated on an own-made support enabling rotation in three planes; (2) zoomed in/out manually by video camera to obtain corresponding size; (3) rotated manually to obtain best position according to vertical (glabella-nasion-prosthion), horizontal (ectocanthion-ectocanthion and porion-porion, when available) and anteroposterior (prosthion-porion) planes. | Morphological assessment of landmarks and outlines. |
P26 Both datasets 88.13% | Adobe Photoshop™ CS3 Video system (Wilkinson 2004) | Follow the procedure described in Jayaprakash et al. (2001) | Asses traits indicated in Jayaprakash et al. (2001). |
Participant number and global performance | Software and equipment | Skull-face overlay | Decision making | Observation |
---|---|---|---|---|
P6–P13 68.57–82.35% Male dataset | Adobe Photoshop™ or GNU GIMP. Dennis Slice’s PhotoMatch | Photoshop to resize images. Then, PhotoMatch to manipulate transparency, scaling, and rotation to match as many points as possible. | Special signs: crowding, broken nasal bones, asymmetry, etc. | Do not use the 3D data but just the PM skull photos. Limited orientation and alignment capabilities. |
7.2.2 Results
Participant N. | N | N of P | N of P given | TP (%) | N of N | N of N given | TN (%) | N of N | N of P given | FP (%) | N of P | N of N given | FN (%) |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Baseline | 30 | 5 | 5 | 100.00 | 25 | 25 | 100.00 | 25 | 0 | 0.00 | 5 | 0 | 0.00 |
P1 | 30 | 5 | 3 | 60.00 | 25 | 21 | 84.00 | 25 | 4 | 16.00 | 5 | 2 | 40.00 |
P2 | 30 | 5 | 4 | 80.00 | 25 | 24 | 96.00 | 25 | 1 | 4.00 | 5 | 1 | 20.00 |
P4 | 30 | 5 | 4 | 80.00 | 25 | 24 | 96.00 | 25 | 1 | 4.00 | 5 | 1 | 20.00 |
P5 | 30 | 5 | 4 | 80.00 | 25 | 23 | 92.00 | 25 | 2 | 8.00 | 5 | 1 | 20.00 |
P6 | 4 | 2 | 1 | 50.00 | 2 | 2 | 100.00 | 2 | 0 | 0.00 | 2 | 1 | 50.00 |
P7 | 30 | 5 | 2 | 40.00 | 25 | 21 | 84.00 | 25 | 4 | 16.00 | 5 | 3 | 60.00 |
P8 | 30 | 5 | 1 | 20.00 | 25 | 21 | 84.00 | 25 | 4 | 16.00 | 5 | 4 | 80.00 |
P9 | 29 | 4 | 1 | 25.00 | 25 | 23 | 92.00 | 25 | 2 | 8.00 | 4 | 3 | 75.00 |
P10 | 30 | 5 | 1 | 20.00 | 25 | 20 | 80.00 | 25 | 5 | 20.00 | 5 | 4 | 80.00 |
P11 | 29 | 5 | 3 | 60.00 | 24 | 17 | 70.83 | 24 | 7 | 29.17 | 5 | 2 | 40.00 |
P12 | 30 | 5 | 1 | 20.00 | 25 | 21 | 84.00 | 25 | 4 | 16.00 | 5 | 4 | 80.00 |
P13 | 30 | 5 | 2 | 40.00 | 25 | 20 | 80.00 | 25 | 5 | 20.00 | 5 | 3 | 60.00 |
P14 | 30 | 5 | 4 | 80.00 | 25 | 22 | 88.00 | 25 | 3 | 12.00 | 5 | 1 | 20.00 |
P15 | 30 | 5 | 3 | 60.00 | 25 | 21 | 84.00 | 25 | 4 | 16.00 | 5 | 2 | 40.00 |
P16 | 20 | 3 | 0 | 0.00 | 17 | 11 | 64.71 | 17 | 6 | 35.29 | 3 | 3 | 100.00 |
P17 | 20 | 5 | 4 | 80.00 | 15 | 11 | 73.33 | 15 | 4 | 26.67 | 5 | 1 | 20.00 |
P18 | 29 | 4 | 3 | 75.00 | 25 | 23 | 92.00 | 25 | 2 | 8.00 | 4 | 1 | 25.00 |
P19 | 26 | 3 | 2 | 66.67 | 23 | 18 | 78.26 | 23 | 5 | 21.74 | 3 | 1 | 33.33 |
P20 | 30 | 5 | 3 | 60.00 | 25 | 21 | 84.00 | 25 | 4 | 16.00 | 5 | 2 | 40.00 |
P21 | 28 | 5 | 4 | 80.00 | 23 | 18 | 78.26 | 23 | 5 | 21.74 | 5 | 1 | 20.00 |
P22 | 30 | 5 | 5 | 100.00 | 25 | 20 | 80.00 | 25 | 5 | 20.00 | 5 | 0 | 0.00 |
P23 | 29 | 4 | 2 | 50.00 | 25 | 14 | 56.00 | 25 | 11 | 44.00 | 4 | 2 | 50.00 |
P24 | 30 | 5 | 0 | 0.00 | 25 | 24 | 96.00 | 25 | 1 | 4.00 | 5 | 5 | 100.00 |
P26 | 29 | 4 | 3 | 75.00 | 25 | 22 | 88.00 | 25 | 3 | 12.00 | 4 | 1 | 25.00 |
P25 | 30 | 5 | 3 | 60.00 | 25 | 20 | 80.00 | 25 | 5 | 20.00 | 5 | 2 | 40.00 |
Participant N. | N | N of P | N of P given | TP (%) | N of N | N of N given | TN (%) | N of N | N of P given | FP (%) | N of P | N of N given | FN (%) |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Baseline | 30 | 5 | 5 | 100.00 | 25 | 25 | 100.00 | 25 | 0 | 0.00 | 5 | 0 | 0.00 |
P1 | 29 | 5 | 2 | 40.00 | 24 | 20 | 83.33 | 24 | 4 | 16.67 | 5 | 3 | 60.00 |
P2 | 30 | 5 | 4 | 80.00 | 25 | 24 | 96.00 | 25 | 1 | 4.00 | 5 | 1 | 20.00 |
P3 | 30 | 5 | 3 | 60.00 | 25 | 24 | 96.00 | 25 | 1 | 4.00 | 5 | 2 | 40.00 |
P5 | 28 | 4 | 1 | 25.00 | 24 | 18 | 75.00 | 24 | 6 | 25.00 | 4 | 3 | 75.00 |
P14 | 30 | 5 | 1 | 20.00 | 25 | 23 | 92.00 | 25 | 2 | 8.00 | 5 | 4 | 80.00 |
P15 | 29 | 5 | 1 | 20.00 | 24 | 20 | 83.33 | 24 | 4 | 16.67 | 5 | 4 | 80.00 |
P16 | 28 | 5 | 3 | 60.00 | 23 | 19 | 82.61 | 23 | 4 | 17.39 | 5 | 2 | 40.00 |
P17 | 30 | 5 | 1 | 20.00 | 25 | 22 | 88.00 | 25 | 3 | 12.00 | 5 | 4 | 80.00 |
P18 | 29 | 4 | 2 | 50.00 | 25 | 22 | 88.00 | 25 | 3 | 12.00 | 4 | 2 | 50.00 |
P19 | 27 | 4 | 3 | 75.00 | 23 | 17 | 73.91 | 23 | 6 | 26.09 | 4 | 1 | 25.00 |
P20 | 30 | 5 | 3 | 60.00 | 25 | 22 | 88.00 | 25 | 3 | 12.00 | 5 | 2 | 40.00 |
P22 | 30 | 5 | 3 | 60.00 | 25 | 18 | 72.00 | 25 | 7 | 28.00 | 5 | 2 | 40.00 |
P26 | 30 | 5 | 3 | 60.00 | 25 | 24 | 96.00 | 25 | 1 | 4.00 | 5 | 2 | 40.00 |
P25 | 30 | 5 | 4 | 80.00 | 25 | 23 | 92.00 | 25 | 2 | 8.00 | 5 | 1 | 20.00 |
Participant N. | N | N of P | N of P given | TP (%) | N of N | N of N given | TN (%) | N of N | N of P given | FP (%) | N of P | N of N given | FN (%) |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Baseline | 30 | 5 | 5 | 100.00 | 25 | 25 | 100.00 | 25 | 0 | 0.00 | 5 | 0 | 0.00 |
P1 | 59 | 10 | 5 | 50.00 | 49 | 41 | 83.67 | 49 | 8 | 16.33 | 10 | 5 | 50.00 |
P2 | 60 | 10 | 8 | 80.00 | 50 | 48 | 96.00 | 50 | 2 | 4.00 | 10 | 2 | 20.00 |
P5 | 58 | 9 | 5 | 55.56 | 49 | 41 | 83.67 | 49 | 8 | 16.33 | 9 | 4 | 44.44 |
P14 | 60 | 10 | 5 | 50.00 | 50 | 45 | 90.00 | 50 | 5 | 10.00 | 10 | 5 | 50.00 |
P15 | 59 | 10 | 4 | 40.00 | 49 | 41 | 83.67 | 49 | 8 | 16.33 | 10 | 6 | 60.00 |
P16 | 48 | 8 | 3 | 37.50 | 40 | 30 | 75.00 | 40 | 10 | 25.00 | 8 | 5 | 62.50 |
P17 | 50 | 10 | 5 | 50.00 | 40 | 33 | 82.50 | 40 | 7 | 17.50 | 10 | 5 | 50.00 |
P18 | 58 | 8 | 5 | 62.50 | 50 | 45 | 90.00 | 50 | 5 | 10.00 | 8 | 3 | 37.50 |
P19 | 53 | 7 | 4 | 57.14 | 46 | 40 | 86.96 | 46 | 6 | 13.04 | 7 | 3 | 42.86 |
P20 | 60 | 10 | 6 | 60.00 | 50 | 43 | 86.00 | 50 | 7 | 14.00 | 10 | 4 | 40.00 |
P22 | 60 | 10 | 8 | 80.00 | 50 | 38 | 76.00 | 50 | 12 | 24.00 | 10 | 2 | 20.00 |
P25 | 60 | 10 | 7 | 70.00 | 50 | 43 | 86.00 | 50 | 7 | 14.00 | 10 | 3 | 30.00 |
P26 | 59 | 9 | 6 | 66.67 | 50 | 46 | 92.00 | 50 | 4 | 8.00 | 9 | 3 | 33.33 |
Participants | Sample | Correct decisions | Ground truth | Decision | Decision (%) | ||
---|---|---|---|---|---|---|---|
Positive | Negative | Positive | Negative | ||||
All | Global | 78.99% | Positive | 100 | 90 | 52.63 | 47.37 |
Negative | 152 | 810 | 15.80 | 84.20 | |||
1 | Global | 81.88% | Positive | 14 | 9 | 60.87 | 39.13 |
Negative | 16 | 99 | 13.91 | 86.09 | |||
2 | Global | 74.32% | Positive | 15 | 33 | 31.25 | 68.75 |
Negative | 43 | 205 | 17.34 | 82.66 | |||
3 | Global | 84.68% | Positive | 22 | 12 | 64.71 | 35.29 |
Negative | 20 | 155 | 11.43 | 88.57 | |||
4 | Global | 75.00% | Positive | 18 | 14 | 56.25 | 43.75 |
Negative | 34 | 126 | 21.25 | 78.75 | |||
5 | Global | 80.75% | Positive | 31 | 22 | 58.49 | 41.51 |
Negative | 39 | 225 | 14.77 | 85.23 |
Participants N. | Tech. approach | Correct decision | Ground truth | Decision | Decision (%) | ||
---|---|---|---|---|---|---|---|
Positive | Negative | Positive | Negative | ||||
P2, P19 | CAs3DS | 88.49% | Positive | 12 | 5 | 70.59 | 29.41 |
Negative | 8 | 88 | 8.33 | 91.67 | |||
P1, P3, P4, P5, P14, P15, P16, P17, P18, P20, P21, P22, P23 | CAm3DS | 78.90% | Positive | 62 | 47 | 56.88 | 43.12 |
Negative | 92 | 458 | 16.73 | 83.27 | |||
P24, P25, P26 | CAmVS | 84.56% | Positive | 13 | 11 | 54.17 | 45.83 |
Negative | 12 | 133 | 8.28 | 91.72 | |||
P6, P7, P8, P9, P10, P11, P12, P13 | CAmPS | 74.06% | Positive | 12 | 24 | 33.33 | 66.67 |
Negative | 31 | 145 | 17.61 | 82.39 |
7.2.2.1 Set of Criteria for Assessing the Skull-Face Overlay Relationship
No. | Criteria | View | Do you employ it? |
---|---|---|---|
Group 1 | Superimposition of the following marking lines (face–skull): | Frontal (F) | – |
1.1 | excanthion-excanthion (A) – ectocanthion-ectoconchion (A′) Ectocanthion line | F | |
1.2 | glabella-gnathion (B) – glabella-gnathion (B′) Frontal central line | F | |
1.3 | superciliary-superciliary (C) – superciliary-superciliary (C′) Supraciliary line | F | |
1.4 | horizontal line at subnasal (D) – horizontal line at nasospinal (D′) Subnasal line | F | |
1.5 | cheilion-cheilion (E) – occusal line/horizontal line at stomion (E′) Cheilion line | F | |
1.6 | horizontal line at gnathion (H) – horizontal line at gnathion (H′) Gnathion line | F | |
1.7 | endocanthion-cheilion (G) – entocanthion-caninion (G′) [right] Entocanthion vertical line | F | |
1.8 | endocanthion-cheilion (G) – entocanthion-caninion (G′) [left] Entocanthion vertical line | F |
No. | Criteria | View | Do you employ it? |
---|---|---|---|
Group 2 | Overall consistency of the facial outline and facial soft tissue thickness at the following pair of homologous points (skull-face): | Frontal (F)/Lateral-Oblique (L-O) | – |
2.1 | gnathion – menton | F | |
2.2 | gnathion – menton | L-O | |
2.3 | nasion – nasion | L-O | |
2.4 | glabella – glabella | L-O | |
2.5 | subespinale – subnasale | L-O | |
2.6 | pogonion – pogonion | L-O | |
2.7 | rhinion – rhinion | L-O | |
2.8 | gonion – gonion | F | |
2.9 | zygion – zygion | F | |
2.10 | trichion – trichion | L-O | |
2.11 | The minimal tissue thickness all along the contour is considered from the point of view of its symmetry by side for evaluating the match as acceptable. | ||
2.12 | Ant lacrimal crest – medial canthus | F/O | |
2.13 | Prosthion – supra-labiale | F/L-O | |
2.14 | Alare – Alare | F/L-O | |
2.15 | Gonion – gonion | L | |
2.16 | Zygomaxilare – malare | L | |
2.17 | Whitnall’s tubercle – lateral canthus | F/L-O | |
2.18 | occlusion mid-incisors – stomion | F/L-O | |
2.19 | porion – tragion | L-O | |
2.20 | crista conchalis – supra-alare | F/L-O | |
2.21 | intercanine distance (75%) – chelion | F | |
2.22 | eyeball position – pupilare | F | |
2.23 | supraorbitale – sag eyebrow | F | |
2.24 | two tangents nasal – pronasale | F/L-O | |
2.25 | 1st premolar/canine radiating line – chelion | F/L-O | |
2.26 | mastoidale – subaurale | L-O | |
2.27 | infraorbital foramen – chelion | F/L-O |
No. | Criteria | View | Do you employ it? |
---|---|---|---|
Group 3 | Overall consistency of the bony and facial outlines/morphological curves: | Frontal(F)/Lateral-Oblique(L-O) | – |
3.1 | The outline of the frontal bone follows the forehead outline. | L-O | |
3.2 | The skull and head height is similar (account for variation in soft tissue and distortion in the perception created by presence of hair). | L-O | |
3.3 | The width of the cranium fills forehead area of the face. | F | |
3.4 | The length of the skull from menton to bregma fits within the face. | F | |
3.5 | The lateral line of the zygomatic bone matches the outline of cheek. | F | |
3.6 | The chin outline is consistent with the mental outline. | L-O | |
3.7 | The gonial outline follows the outline of jaw angle. | L-O | |
3.8 | The outline of frontal process of the zygomatic bone can be aligned with the process seen in the face. The outline of the zygomatic arch can be fitted between the skull and the face. (This criterion is more easily appreciated in individuals with minimal soft tissue thickness). | L-O | |
3.9 | The arcus supraciliariaris follows supraorbital margin. | L-O | |
3.10 | The temporal line is consistent with the outline of the forehead (Sometimes the temporal line cannot be distinguished). | F | |
3.11 | The outline of the face and the outline of the skull all along the contour follow each other maintaining symmetrical flow by side. | F | |
3.12 | The outline of the nose in the face represented by shade distribution follows the outline of the nasal bone in the skull maintaining symmetrical flow by side. | F | |
3.13 | The asymmetries in the facial organs especially the nose reveal consistency with the asymmetries in the organs of the skull including the nasal structures. These include 1. Asymmetries in the nasal area including the nasal bone, piriform aperture, and nasal spine 2. Asymmetries in the zygomatic area, especially the extent of protrusion of the arch 3. Asymmetries in the occlusal line caused by protruding or overriding anterior dentition reflected as asymmetries in the corresponding part of the lip closure line 4. Asymmetries in the gonia | F | |
3.14 | The arcus supraciliariaris follows supraorbital margin. | F | |
3.15 | The outline of the of the nasal bones follows the outline of the nose in the skull with minimal tissue thickness allowance | L | |
3.16 | The sagittal outline of the nasal cartilage is the mirror image of the contour of the pyriform aperture, relative to Line №1 passing through the rhinion point (1) and parallel to Line №2 joining the nasion(2) and the prostion (3) anthropometric points. Line№1 splits the entire nasal cartilage into two symmetric mirrored halves: the protruding part of the nose cartilage is the mirror image of the cartilage filling the pyriform aperture of the cranium. | ||
3.17 | Lateral nasal bulges. | F | |
3.18 | Oblique line of the mandible. | F | |
3.19 | Dental information (bony to bony consistency). | F/L-O |
No. | Criteria | View | Do you employ it? |
---|---|---|---|
Group 4 | Overall consistency positional relationships between the skull and face: | Frontal (F)/Lateral-Oblique (L-O) | – |
4.1 | The prosthion lies posterior to the anterior edge of the upper lip. The occlusal and the lip closure line are consistent. | F | |
4.2 | The lateral angle of the eye lies within the lateral wall of the orbit. | L-O | |
4.3 | The lateral orbital margin at the Whinall’s turbercle matches or approximates the of the ectocathionposition | L-O | |
4.4 | The width and height of piriform aperture lie within the borders of the nose. | F | |
4.5 | The stomion lies at the central incisors (at the occusal line). | L-O | |
4.6 | The lateral margin of the piriform aperture matches or approximates the alare. | L-O | |
4.7 | The stomion lies at the central incisors (incisal margin of the upper incisors). | F | |
4.8 | The nasion is higher than the nasal root. | L-O | |
4.9 | The prosthion lies posterior to the anterior edge of the upper lip. The occlusal and the lip closure line are consistent. | L-O | |
4.10 | The medial margin of orbit aligns and superimposes with the endocanthion. | F | |
4.11 | The nasion is higher than the nasal root. | F | |
4.12 | The Whitnall’s tubercle aligns with the ectochantus on the horizontal plane and ,vertically, the ectochantus lies medial to the tubercle. The orbital width is consistent with the eye-slit width. | F | |
4.13 | The chelion lies between the canine and the first premolar (at the occusal line). | F | |
4.14 | The porion aligns just posterior to the tragus, slightly inferior to the crus of the helix. | L-O | |
4.15 | The lower margin of piriform aperture matches the subnasale. | L-O | |
4.16 | The eyebrow generally follows the upper edge of the orbit over the medial two-thirds. At lateral superior one-third of the orbit, the eyebrow continues horizontally as the orbital rim begins to curve inferiorly. | F | |
4.17 | The external auditory meatus opening lies medial to the tragus of the ear. (Place a projecting marker at the ear canal to assess this criterion more easily). | F | |
4.18 | The chelion lies between the canine and the first premolar (at the occusal line). | L-O | |
4.19 | The anterior nasal spine lies posterior to the base of the nose near the most posterior portion of the lateral septal cartilage. | L-O | |
4.20 | Gonial flare in the skull and the postero-lateral jaw angle outline in the face. | L-O | |
4.21 | Gonial flare in the skull and the postero-lateral jaw prominence in the face. | F |
7.3 Study on the Criteria Assessing Skull-Face Correspondence in Craniofacial Superimposition
7.3.1 Experimental Study
7.3.1.1 Data Analysis
-
Cases with decision (CD): the cases in which the expert’s decision is different from 0 (not undetermined)
-
Expert proficiency: the proportion of cases with decision in which the expert evaluated the status of the case correctly
7.3.1.2 Correlation Between Two Variables
-
Criterion weighted correlation with ground truth: same as the correlation with ground truth, except that the correlation coefficient associated with each expert is weighted according to his proficiency.
-
Criterion variability: it is computed as the mean standard deviation of the criterion evaluation over the same case. It aims to assess the subjectivity of a criterion.
7.3.1.3 Linear Regression
7.3.2 Results
Forensic expert | Proficiency | Cases with decision |
---|---|---|
F1 | 0.62 | 0.96 |
F2 | 0.58 | 1.00 |
F3 | 0.58 | 0.83 |
F4 | 0.50 | 1.00 |
F5 | 0.54 | 0.79 |
F6 | 0.38 | 0.96 |
F7 | 0.62 | 1.00 |
F8 | 0.46 | 0.96 |
F9 | 0.42 | 0.96 |
F10 | 0.42 | 0.83 |
F11 | 0.48 | 0.96 |
F12 | 0.62 | 1.00 |
F13 | 0.54 | 0.83 |
F14 | 0.50 | 0.92 |
F15 | 0.58 | 1.00 |
F16 | 0.67 | 1.00 |
F17 | 0.67 | 1.00 |
F18 | 0.46 | 0.88 |
F19 | 0.54 | 1.00 |
F20 | 0.58 | 0.92 |
F21 | 0.67 | 0.96 |
F22 | 0.67 | 0.88 |
F23 | 0.75 | 1.00 |
F24 | 0.50 | 1.00 |
F25 | 0.50 | 0.88 |
F26 | 0.42 | 0.83 |
F27 | 0.67 | 1.00 |
F28 | 0.42 | 0.96 |
F29 | 0.50 | 1.00 |
F30 | 0.62 | 1.00 |
F31 | 0.75 | 0.96 |
F32 | 0.46 | 1.00 |
F33 | 0.62 | 1.00 |
F34 | 0.67 | 0.92 |
F35 | 0.54 | 0.92 |
F36 | 0.54 | 0.75 |
F37 | 0.58 | 1.00 |
-
The absence of a complete cranium.
-
The quality of some 3D models, which in some cases present noisy parts and artifacts.
-
The materials given to the participants do not include the 3D skull models but just a projection on a 2D plane.
-
The isolation of the decision-making stage given an SFO.
Forensic expert | Correlation coefficient | p-Value |
---|---|---|
F1 | 0.29 | 0.17 |
F2 | 0.21 | 0.33 |
F3 | 0.32 | 0.13 |
F4 | 0.21 | 0.33 |
F5 | 0.33 | 0.11 |
F6 | −0.07 | 0.75 |
F7 | 0.03 | 0.91 |
F8 | −0.08 | 0.71 |
F9 | −0.10 | 0.64 |
F10 | 0.13 | 0.53 |
F11 | 0.24 | 0.27 |
F12 | 0.26 | 0.21 |
F13 | 0.26 | 0.21 |
F14 | 0.10 | 0.63 |
F15 | 0.30 | 0.15 |
F16 | 0.38 | 0.07 |
F17 | 0.34 | 0.11 |
F18 | 0.25 | 0.23 |
F19 | 0.22 | 0.31 |
F20 | 0.61 |
0.00
|
F21 | 0.45 |
0.03
|
F22 | 0.52 |
0.01
|
F23 | 0.49 |
0.01
|
F24 | 0.10 | 0.63 |
F25 | 0.23 | 0.29 |
F26 | −0.03 | 0.88 |
F27 | 0.21 | 0.33 |
F28 | −0.22 | 0.30 |
F29 | −0.05 | 0.81 |
F30 | 0.27 | 0.21 |
F31 | 0.54 |
0.01
|
F32 | −0.30 | 0.15 |
F33 | 0.29 | 0.17 |
F34 | 0.50 |
0.01
|
F35 | 0.30 | 0.16 |
F36 | 0.40 | 0.05 |
F37 | 0.13 | 0.54 |
Criterion | All | Frontal | Lateral | Criterion | All | Frontal | Lateral | ||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
UsageC | UsageP | UsageC | UsageP | UsageC | UsageP | UsageC | UsageP | UsageC | UsageP | UsageC | UsageP | ||
G2.14 | 0.72 | 0.83 | 0.78 | 0.83 | 0.67 | 0.78 | G2.16 | 0.41 | 0.64 | 0.23 | 0.32 | 0.58 | 0.64 |
G3.6 | 0.67 | 0.97 | 0.46 | 0.64 | 0.88 | 0.97 | G2.15 | 0.40 | 0.67 | 0.25 | 0.37 | 0.56 | 0.67 |
G2.13 | 0.62 | 0.75 | 0.64 | 0.70 | 0.60 | 0.75 | G4.20 | 0.40 | 0.73 | 0.25 | 0.35 | 0.56 | 0.70 |
G2.3 | 0.59 | 0.94 | 0.36 | 0.48 | 0.88 | 0.94 | G3.8 | 0.39 | 0.70 | 0.26 | 0.43 | 0.52 | 0.62 |
G2.5 | 0.58 | 0.86 | 0.37 | 0.45 | 0.81 | 0.86 | G4.7 | 0.39 | 0.62 | 0.53 | 0.56 | 0.25 | 0.40 |
G4.16 | 0.57 | 0.86 | 0.84 | 0.86 | 0.32 | 0.54 | G2.11 | 0.39 | 0.45 | 0.45 | 0.45 | 0.33 | 0.37 |
G3.15 | 0.56 | 0.91 | 0.27 | 0.43 | 0.88 | 0.91 | G1.6 | 0.38 | 0.62 | 0.59 | 0.62 | 0.18 | 0.29 |
G4.2 | 0.56 | 0.89 | 0.43 | 0.51 | 0.70 | 0.86 | G2.7 | 0.38 | 0.67 | 0.12 | 0.18 | 0.64 | 0.67 |
G2.1 | 0.55 | 0.83 | 0.80 | 0.83 | 0.33 | 0.43 | G4.11 | 0.38 | 0.73 | 0.49 | 0.73 | 0.26 | 0.35 |
G4.1 | 0.54 | 0.94 | 0.71 | 0.78 | 0.40 | 0.54 | G2.4 | 0.38 | 0.94 | 0.24 | 0.37 | 0.51 | 0.91 |
G3.7 | 0.54 | 0.89 | 0.36 | 0.48 | 0.72 | 0.89 | G4.12 | 0.37 | 0.59 | 0.56 | 0.59 | 0.19 | 0.32 |
G3.12 | 0.53 | 0.91 | 0.83 | 0.86 | 0.25 | 0.43 | G4.13 | 0.37 | 0.67 | 0.56 | 0.64 | 0.18 | 0.29 |
G2.9 | 0.52 | 0.89 | 0.85 | 0.89 | 0.20 | 0.35 | G2.12 | 0.36 | 0.51 | 0.50 | 0.51 | 0.22 | 0.43 |
G3.9 | 0.52 | 0.75 | 0.41 | 0.48 | 0.65 | 0.75 | G4.19 | 0.36 | 0.70 | 0.06 | 0.16 | 0.66 | 0.70 |
G3.11 | 0.52 | 0.86 | 0.84 | 0.86 | 0.20 | 0.43 | G4.21 | 0.36 | 0.62 | 0.58 | 0.62 | 0.11 | 0.35 |
G4.4 | 0.52 | 0.78 | 0.78 | 0.78 | 0.26 | 0.43 | G4.18 | 0.35 | 0.59 | 0.22 | 0.32 | 0.43 | 0.56 |
G3.5 | 0.51 | 0.91 | 0.86 | 0.91 | 0.18 | 0.45 | G2.16 | 0.32 | 0.62 | 0.11 | 0.21 | 0.50 | 0.59 |
G4.15 | 0.51 | 0.73 | 0.42 | 0.54 | 0.62 | 0.70 | G2.23 | 0.31 | 0.51 | 0.48 | 0.51 | 0.11 | 0.27 |
G4.8 | 0.51 | 0.81 | 0.32 | 0.40 | 0.71 | 0.78 | G1.7 | 0.30 | 0.56 | 0.48 | 0.56 | 0.08 | 0.24 |
G1.1 | 0.50 | 0.86 | 0.86 | 0.86 | 0.16 | 0.40 | G3.17 | 0.28 | 0.48 | 0.40 | 0.45 | 0.15 | 0.27 |
G3.14 | 0.50 | 0.73 | 0.70 | 0.73 | 0.29 | 0.43 | G1.8 | 0.28 | 0.56 | 0.48 | 0.54 | 0.06 | 0.18 |
G2.2 | 0.48 | 0.75 | 0.27 | 0.32 | 0.69 | 0.75 | G1.2 | 0.27 | 0.70 | 0.40 | 0.70 | 0.10 | 0.29 |
G2.8 | 0.47 | 0.78 | 0.69 | 0.78 | 0.25 | 0.43 | G2.25 | 0.25 | 0.32 | 0.27 | 0.32 | 0.21 | 0.27 |
G2.18 | 0.46 | 0.56 | 0.47 | 0.54 | 0.46 | 0.54 | G2.27 | 0.24 | 0.27 | ||||
G4.3 | 0.46 | 0.70 | 0.42 | 0.48 | 0.50 | 0.59 | G2.22 | 0.24 | 0.45 | 0.44 | 0.45 | 0.03 | 0.10 |
G4.6 | 0.46 | 0.73 | 0.33 | 0.484 | 0.59 | 0.64 | G2.24 | 0.24 | 0.27 | ||||
G4.10 | 0.44 | 0.78 | 0.75 | 0.78 | 0.13 | 0.40 | G2.21 | 0.21 | 0.37 | 0.33 | 0.37 | 0.05 | 0.16 |
G4.9 | 0.44 | 0.78 | 0.22 | 0.29 | 0.65 | 0.75 | G3.16 | 0.19 | 0.35 | 0.12 | 0.16 | 0.24 | 0.29 |
G4.5 | 0.44 | 0.73 | 0.33 | 0.48 | 0.54 | 0.73 | G3.10 | 0.18 | 0.54 | 0.26 | 0.54 | 0.09 | 0.32 |
G1.4 | 0.43 | 0.67 | 0.64 | 0.67 | 0.22 | 0.35 | G3.19 | 0.18 | 0.48 | 0.18 | 0.48 | 0.17 | 0.48 |
G1.5 | 0.42 | 0.73 | 0.66 | 0.73 | 0.18 | 0.35 | G3.1 | 0.18 | 0.75 | 0.09 | 0.13 | 0.26 | 0.75 |
G3.13 | 0.42 | 0.73 | 0.70 | 0.73 | 0.13 | 0.29 | G2.20 | 0.17 | 0.24 | ||||
G2.17 | 0.41 | 0.45 | 0.45 | 0.45 | 0.37 | 0.40 | G4.14 | 0.16 | 0.51 | 0.03 | 0.08 | 0.24 | 0.51 |
G1.3 | 0.41 | 0.75 | 0.68 | 0.73 | 0.14 | 0.35 | G2.19 | 0.13 | 0.48 | 0.02 | 0.10 | 0.22 | 0.48 |
G3.3 | 0.41 | 0.81 | 0.73 | 0.81 | 0.09 | 0.29 | G4.17 | 0.15 | 0.37 | ||||
G3.18 | 0.41 | 0.67 | 0.56 | 0.62 | 0.26 | 0.43 |
Forensic expert | Criterion | Correlation coefficient | p-Value |
---|---|---|---|
F3 | G2.9 | 0.48 | 0.02 |
F3 | G2.14 | 0.51 | 0.01 |
F4 | G1.4 | 0.43 | 0.03 |
F4 | G3.18 | 0.49 | 0.01 |
F7 | G2.18 | 0.41 | 0.04 |
F11 | G3.12 | 0.47 | 0.02 |
F11 | G3.13 | 0.44 | 0.03 |
F11 | G3.15 | 0.65 | 0.00 |
F16 | G1.2 | 0.42 | 0.03 |
F16 | G1.5 | 0.41 | 0.04 |
F16 | G1.8 | 0.52 | 0.00 |
F16 | G2.3 | 0.49 | 0.01 |
F16 | G2.6 | 0.52 | 0.00 |
F16 | G2.14 | 0.56 | 0.00 |
F16 | G3.6 | 0.46 | 0.02 |
F16 | G4.14 | 0.50 | 0.01 |
F19 | G2.8 | 0.47 | 0.02 |
F19 | G3.7 | 0.43 | 0.03 |
F19 | G4.15 | 0.50 | 0.01 |
F21 | G3.19 | 0.41 | 0.04 |
F30 | G3.18 | 0.48 | 0.01 |
Forensic expert | Criterion | F value | p-Value |
---|---|---|---|
F3 | G2.9 | 5.97 | 0.02 |
F3 | G2.14 | 7.53 | 0.01 |
F4 | G1.4 | 5.00 | 0.03 |
F4 | G3.18 | 6.66 | 0.01 |
F7 | G2.18 | 5.96 | 0.02 |
F11 | G3.12 | 4.62 | 0.04 |
F11 | G3.15 | 11.56 | 0.00 |
F11 | G4.7 | 4.10 | 0.05 |
F16 | G1.5 | 4.66 | 0.04 |
F16 | G1.8 | 7.97 | 0.00 |
F16 | G2.3 | 6.82 | 0.01 |
F16 | G2.6 | 7.97 | 0.00 |
F16 | G2.14 | 8.00 | 0.00 |
F16 | G3.6 | 4.23 | 0.05 |
F16 | G3.12 | 4.45 | 0.04 |
F16 | G4.14 | 7.61 | 0.01 |
F18 | G4.3 | 4.22 | 0.05 |
F19 | G2.8 | 5.03 | 0.03 |
F19 | G2.17 | 4.50 | 0.04 |
F19 | G2.18 | 4.14 | 0.05 |
F19 | G4.15 | 5.76 | 0.02 |
F21 | G3.19 | 5.55 | 0.02 |
F23 | G2.11 | 4.00 | 0.05 |
F30 | G3.18 | 7.48 | 0.01 |
Criterion | Mean standard deviation | Criterion | Mean standard deviation |
---|---|---|---|
G2.20 | 0.85 | G3.16 | 1.12 |
G3.10 | 0.87 | G2.12 | 1.12 |
G3.1 | 0.92 | G2.17 | 1.12 |
G1.5 | 0.92 | G2.14 | 1.12 |
G1.8 | 0.93 | G3.11 | 1.14 |
G3.18 | 0.97 | G3.9 | 1.14 |
G1.3 | 0.97 | G4.12 | 1.14 |
G2.19 | 0.97 | G2.3 | 1.14 |
G2.24 | 1.01 | G3.14 | 1.14 |
G1.2 | 1.02 | G2.15 | 1.14 |
G2.16 | 1.02 | G4.20 | 1.14 |
G2.23 | 1.03 | G3.6 | 1.15 |
G4.19 | 1.03 | G2.22 | 1.15 |
G1.7 | 1.04 | G3.8 | 1.15 |
G2.25 | 1.05 | G4.4 | 1.15 |
G4.10 | 1.06 | G4.6 | 1.15 |
G1.6 | 1.07 | G1.1 | 1.17 |
G2.21 | 1.07 | G2.11 | 1.17 |
G2.9 | 1.07 | G2.5 | 1.18 |
G2.27 | 1.08 | G4.15 | 1.19 |
G4.13 | 1.10 | G4.8 | 1.19 |
G2.7 | 1.10 | G3.5 | 1.19 |
G2.18 | 1.10 | G4.3 | 1.19 |
G2.8 | 1.10 | G4.16 | 1.19 |
G3.7 | 1.10 | G4.2 | 1.20 |
G1.4 | 1.10 | G2.1 | 1.20 |
G4.5 | 1.10 | G4.18 | 1.21 |
G4.14 | 1.10 | G3.19 | 1.21 |
G2.4 | 1.11 | G4.11 | 1.21 |
G4.1 | 1.11 | G3.13 | 1.21 |
G3.3 | 1.11 | G32.2 | 1.23 |
G4.21 | 1.11 | G3.12 | 1.23 |
G2.13 | 1.11 | G3.15 | 1.26 |
G4.9 | 1.11 | G2.6 | 1.31 |