Forensic Facial Reconstruction: Soft Tissue Thickness from Medical Record Computed Tomography Images




Simpson, Ashley

Journal Title

Journal ISSN

Volume Title



Unidentified human remains are examined using a multitude of forensic techniques (i.e., DNA and fingerprinting) to identify the individual. The last-resort is facial reconstruction, which aims to produce a likeness of the individual that can be recognized by someone familiar with the person (Wilkinson C. , Forensic Facial Reconstruction, 2004). Facial soft tissue thickness (FSTT) datasets allow forensic artists to transform the unidentified skull into a representation of the individual’s face. Over the years, several methodologies have been studied and recommended to create these datasets (needle puncture, ultrasound, CT, MRI). This paper assesses the use of medical record computed tomography (CT) images to improve the accuracy of forensic facial reconstructions. The vastness of available data in medical records can improve the specificity of each forensic reconstruction. Additionally, using medical records can provide geographical region-specific FSTT values. Region-specific values will be valuable as more cultures and ethnicities migrate and connect. This paper randomly selected 30 medical records of individuals ages 21-38 from three different ancestral groups: Caucasian, African American, and Hispanic. This study grouped subjects by sex and race for statistical comparison. FSTT was collected at eight facial landmarks using a DICOM viewer called PACS. The tissue depths measured were analyzed and compared with previous CT datasets. ANOVA results showed two statistically significant (p<0.05) landmarks at the infraorbital and zygomatic arch. Due to time restrictions, the sample size used for this study was very small resulting in a large percent error. Understandably the data from this study alone is not helpful to the field of forensic facial reconstruction, but the future potential described in this paper is promising. One way to increase the accuracy of facial reconstructions is to identify groups of similar morphology. The more we expand the dataset across individuals of various ages and races, the more narrowly the groups can be defined. This paper shows that FSTT collected from medical records is a valuable source of data for facial reconstruction datasets, but the risk for error is significant. Therefore, an automated method of data extraction that minimizes the risk for error would be a viable future project.



Forensic facial tissue


Simpson, A. P. (2019). Forensic Facial Reconstruction: Soft Tissue Thickness from Medical Record Computed Tomography Images.