Elsevier

Forensic Science International

Volume 261, April 2016, Pages 164.e1-164.e8
Forensic Science International

Forensic Anthropology Population Data
Three-dimensional prediction of the human eyeball and canthi for craniofacial reconstruction using cone-beam computed tomography

https://doi.org/10.1016/j.forsciint.2016.01.031Get rights and content

Highlights

  • Regression models predicting positions of the eyeball and canthi could be generated.

  • Orbital measurements were strong predictors of the eyeball and canthi positions.

  • Intercanine width was a predictor of the mediolateral position of the eyeball.

  • CBCT data can be useful in predicting the positions of the eyeball and canthi.

Abstract

An anatomical relationship between the hard and soft tissues of the face is mandatory for facial reconstruction. The purpose of this study was to investigate the positions of the eyeball and canthi three-dimensionally from the relationships between the facial hard and soft tissues using cone-beam computed tomography (CBCT). CBCT scan data of 100 living subjects were used to obtain the measurements of facial hard and soft tissues. Stepwise multiple regression analyses were carried out using the hard tissue measurements in the orbit, nasal bone, nasal cavity and maxillary canine to predict the most probable positions of the eyeball and canthi within the orbit. Orbital width, orbital height, and orbital depth were strong predictors of the eyeball and canthi position. Intercanine width was also a predictor of the mediolateral position of the eyeball. Statistically significant regression models for the positions of the eyeball and canthi could be derived from the measurements of orbit and maxillary canine. These results suggest that CBCT data can be useful in predicting the positions of the eyeball and canthi three-dimensionally.

Introduction

Craniofacial reconstruction is the process of recreating the face of an individual from skeletal remains through forensic science or anthropology for the purpose of assisting in facial identification [1]. There has been substantial research quantifying the relationship between the skull and the overlying soft tissues of the face with the aim of facilitating facial reconstruction [2]. An accurate anatomical relationship between the bony and soft tissues of the face is essential for facial reconstruction and anthropological research concerned with the prediction of faces [3].

There have been several studies about the placement of the eyeball or canthi [3], [4], [5], [6], cadaver-based studies [3], [4] for instance, can provide the empirical evidence of the positions of the eyeball and canthi with surrounding hard tissue structures. However, these studies can have limitations due to skin shrinkage in the cadavers or the use of only elderly cadaver samples. Wilkinson and Mautner [5] used magnetic resonance image (MRI) for measuring the eyeball protrusion and orbital depth in their study. MRI study provides a clear image of soft tissue structure containing fat and water. However, this study was limited to the only anteroposterior position of the eyeball [5].

Guyomarc’h et al. [6] studied three-dimensional (3D) placement of the eyeball using computed tomography (CT) images in which they suggested a probable position of the eyeball in the orbit. While the relationship of hard and soft structures can be identified three-dimensionally using the CT data, one drawback of a multi-slice CT is that the images are obtained with the subject in a supine position, and not in a natural posture. The eyeball position in a supine position can be different from a seated or an upright posture. In a study regarding comparison in facial soft tissue thickness between ultrasound system and multi-slice CT, De Greef et al. [7] stated that measurement differences were attributed to a difference in the subject's position, in other words, a supine position during CT scan.

Recently-developed cone-beam CT (CBCT) allows images to be obtained with the subject in an upright position. It provides a natural shape of the facial mask as the subject can be scanned in the seated position [8], [9], [10]. CBCT also produces images that are obtained with the natural head posture of the subject, making it useful for a study about the positions of the eyeball and canthi. In addition, it has been reported that the radiation dose is lower in CBCT scan compared with multi-slice CT [11], [12], [13]. Few studies are available on the 3D positions of the eyeball and canthi using CBCT images. The purposes of this study were to investigate the hard and soft tissue spatial relationships using CBCT images, and to generate regression models to predict the eyeball and canthi positions three-dimensionally.

Section snippets

Materials and methods

The subjects were selected from the graduate students at Chonnam National University in Gwangju, Korea. All subjects had Korean ancestry, and none of the subjects had undergone orthodontic treatment. The subjects with an abnormal body mass index, malocclusion, and dentofacial deformity were excluded. The individuals who were over 35 years old were also excluded due to the age-related changes on the face. One hundred students, 50 men (age range: 25.3–31.1 years; mean: 28.6 years; SD: 2.3 years)

Statistical analysis

To assess the reliability of measurements, all measurements were repeated after 2 weeks by the same examiner and in addition, they were also obtained by a second examiner. The intra- and inter-examiner errors were determined using a paired t-test. In addition, the magnitude of measurement errors was assessed by calculating the intraclass correlation coefficient (ICC). The intra- and inter-examiner errors at p < 0.05 were considered to be statistically insignificant. The ICC measurements indicated

Results

The means and standard deviations of the measurements for the eyeball and canthi positions are shown in Table 2. Statistically significant sex differences were present in the orbital width, nasal height, nasal bone length, and nasal bone projection (Table 2).

As a result of correlation analysis of the hard and soft tissue measurements, statistically significant correlations were found (Table 3). The eyeball measurements showed statistically significant correlations with the orbital dimensions,

Discussion

Regression is an approach used for modeling the relationship between a dependent variable and one or more explanatory variables. It is certain that the predictive value of a regression model increases with the use of more variables. A recent study regarding the prediction of the eyeball position used the variables related to the orbit only as the explanatory variables [6]. For our study, to predict a more accurate position of the pupil, medial and lateral canthi, distant structures such as the

Conclusions

Statistically significant regression models for the 3D positions of the eyeball and canthi could be derived from the measurements of the orbit and upper canine. The results of the present study suggest that CBCT data can be useful for predicting the positions of the eyeball and canthi three-dimensionally.

Acknowledgements

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF-2010-0025828).

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (NRF-2014R1A1A1003559).

References (28)

  • S. De Greef et al.

    Semi-automated ultrasound facial soft tissue depth registration: method and validatio

    J. Forensic Sci.

    (2005)
  • H.S. Hwang et al.

    Reproducibility of facial soft tissue thicknesses for craniofacial reconstruction using cone-beam computed tomography images

    J. Forensic Sci.

    (2012)
  • H.S. Hwang et al.

    Facial soft tissue thickness database for craniofacial reconstruction in Korean adults

    J. Forensic Sci.

    (2012)
  • K.M. Lee et al.

    Three-dimensional prediction of the nose for facial reconstruction using cone-beam computed tomography

    Forensic Sci. Int.

    (2014)
  • Cited by (15)

    • A post-mortem computed tomography imaging method for predicting the anteroposterior and superoinferior positions of the Japanese adult eyeball in the orbit

      2022, Forensic Imaging
      Citation Excerpt :

      Most published studies involved Caucasians. Only one study by Kim et al. used Mongoloids as subjects [19], reporting an eye protrusion of 6.6 mm from the plane that included the most superior and inferior orbital points. In the South Korean sample, 54.2% of the OBHs were from the most superior orbital point.

    • Facial Approximation (Reconstruction)

      2022, Encyclopedia of Forensic Sciences: Volume 1-4, Third Edition
    • Facial approximations: Characteristics of the eye in a South African sample

      2018, Forensic Science International
      Citation Excerpt :

      Reports regarding the distance of the exocanthion from the lateral orbital margin are remarkably similar at 4.5 mm [8], 4.7 mm [16] and 5 mm medial to the malar tubercle [23], and are comparable to our findings (5.0 mm). The distance from the medial orbital margin to the endocanthion is more variable: 4.8 mm [8], 9.8 mm [16], approximately 2 mm lateral to the MOM [23] and 4.8 mm (current study). The variation noted in the Korean sample [16] might be due to the presence of epicanthal folds in people of Asian descent.

    • Automated facial recognition and candidate list rank change of computer generated facial approximations generated with multiple eye orb positions

      2016, Forensic Science International
      Citation Excerpt :

      Subject to a degree of artistic privilege, however, is the construction and positioning of the highly variable facial features (e.g., eyes, nose, and lips) [1–7]. A recent development in the field of facial approximation is debate regarding appropriate placement of the eye orbs within the bony eye orbits of the skull [4–8]. Two primary factions exist regarding appropriate eye orb positioning.

    View all citing articles on Scopus
    1

    These authors contributed equally to this work.

    View full text