Forensic Facial Reconstruction (FFR) From CT Scans

3-D Computed Forensic Facial Reconstruction of Computed Tomography (CT) Scanned Skulls Using Human Facial Templates

1. Reconstructing faces derived from CT scanned skulls using the average facial templates and employing computer facial reconstruction software programme.
2. Evaluating the accuracy of facial reconstructions by objective comparison between the reconstructed face and the CT scanned face.
3. Evaluating the accuracy of facial reconstructions by subjective assessment via 2 groups of individuals, experienced and inexperienced in facial images perception.

Study Overview

• Status: Completed
• Conditions: Forensic Facial Reconstruction

Detailed Description

1. Patients Recruitment Patients meeting the inclusion criteria are identified and recruited by the attending clinical staff at a private Diagnostic Radiology Center in Egypt after completing the relevant ethical and regulatory procedures.The study will include a minimum of 20 case studies.
2. Acquisition of CT Scanned Head Images CT images are acquired by a computed tomography scanner.
3. Segmentation (i.e extraction) of a 3D images of the skull from CT Digital Imaging and Communications in Medicine (DICOM) files: via 3D slicer software and Simpleware (CT segmenting software) trial software.
4. Selection of Average Face Templates For each studied case, the facial reconstruction will be conducted using an average Caucasian face template that matches the studied skull, in terms of sex and age group selected.
5. Facial Reconstruction is performed using the patient segmented skull image and its matching average face template, and employing the present computer software program.
6. Segmentation (i.e extraction) of a 3D images of the face surface from CT DICOM files.

7. Objective Assessment of Facial Reconstructions for each case via objective surface superimposition between the extracted 3D facial image and the 3D reconstructed face using Robin's surface viewer software.

   The sum of square differences (SSD) between the superimposed images will be calculated as an indication of the goodness of fit of the reconstructed face with the CT scanned face. The measured differences will then be correlated with the sum of scores given to each reconstructed face by subjective assessment.

8. Subjective Assessment by Volunteers Assessors Volunteer assessors will be approached in an attempt to evaluate the accuracy of the resulting reconstructed images. Two groups of assessors will be recruited; the first assessors will be experts with professional experience in face image identification (n=5), and the second group will comprise of participants inexperienced in forensic facial identification. The latter will be recruited from staff and students of Queen Mary University of London (QMUL) (n=20-30). All participants will be asked to sign a written confidentiality agreement.

   1. Face Pool Test Construction The aim of face pool test in the present study is to assess the ability of observers to identify a target subject from a face pool of similar subjects. The face pool will be formed of computer generated faces, in 5 views (frontal, right lateral and left lateral, right ¾ and left ¾), included the extracted CT face of the target and these faces will be compared to the target facial reconstruction. Each observer will be asked to select one image from the test faces that best matches the target reconstructed face.
   2. Resemblance Test Construction This test aims to compare and assess the similarity between the facial reconstruction and the CT face of a target subject using rating/ranking scale. Each observer will be asked to rate the target reconstructed face from 0 to 10 according to its resemblance to the target CT face , where 0 = no resemblance and 10 = highest resemblance. Facial reconstructions of all cases will be ranked by the total resemblance scores given by all observers to each reconstruction.

9. Statistical analysis of the results via spearman's Rank coefficient to assess and compare:

   1. Facial reconstructions ranks given by the 2 groups of observers to assess the influence of previous experience in face identification on the ability of the observer to assess the accuracy of forensic facial reconstructions.
   2. Facial reconstructions ranks given by subjective and objective assessment methods, thus compare the validity of both methods in assessing the accuracy of forensic facial reconstructions.

• Study Type: Observational
• Enrollment (Actual): 2

Contacts and Locations

Study Locations

• United Kingdom
  • London, United Kingdom, EC1M 6BQ
    • Barts School of Medicine and Dentistry, Queen Mary University of London

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Child; Adult; Older Adult
• Accepts Healthy Volunteers: No

• Sampling Method: Non-Probability Sample

Study Population

1. This study will involve conducting facial reconstruction of 3-D human skull images acquired from head CT scanning of living patients in a private Diagnostic radiology center in Egypt.
2. The resulting facial reconstructions will then be assessed by 2 groups of volunteer participants; experts and non-experts in the field of facial identification respectively.

Description

Inclusion Criteria:

1. Caucasian ethnicity. Caucasian ethnicity is the general physical type of some or all of the populations of Europe, North Africa, the Horn of Africa, Western Asia, Central Asia and South Asia.
2. Above 16 years of age.
3. Able to understand and willing to sign the written Informed Consent Form.
4. Able and willing to follow the protocol requirements.
5. Undergoing CT scanning of the head for medical or clinical purposes.
6. A full CT head series is undertaken.

Exclusion Criteria:

1. Ethnic background other than Caucasian.
2. Less than 16 years old.
3. Unable to understand or refusing to sign the written Informed Consent.
4. Unable or refusing to follow the protocol requirements.
5. Unclear or incomplete head CT scan images.
6. Disfiguring bone or soft tissue deformity or scans with implantable devices (i.e. eye, head and neck, or cardiac).

Study Plan

How is the study designed?

Number of groups / cohorts

1

Cohorts and Interventions

Group / Cohort
Head CT scan; Adult Caucasian patients undergoing head CT scanning.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Time Frame
The percentage of correct identifications and resemblance scores of the target reconstructed faces by human assessors.	3 years

Secondary Outcome Measures

Outcome Measure	Time Frame
The influence of previous experience in the field of facial image recognition and perception on the identification of the reconstructed faces.	3 years

Collaborators and Investigators

• Sponsor: Queen Mary University of London
• Collaborators: Ministry of Higher Education & Scientific Research (Egypt)
• Investigators: Principal Investigator: Dalia AS Abdou, MSc, PhD student, Queen Mary University of London; Principal Investigator: Peter Vanezis, MD, PhD, Professor of Forensic Medical Sciences, Queen Mary University of London; Principal Investigator: Atholl Johnston, PhD, Professor of Clinical Pharmacology, Queen Mary University of London

Study record dates

Study Major Dates

• Study Start (Actual): May 28, 2014
• Primary Completion (Actual): April 30, 2015
• Study Completion (Actual): April 30, 2015

Study Registration Dates

• First Submitted: July 11, 2014
• First Submitted That Met QC Criteria: July 4, 2024
• First Posted (Actual): July 11, 2024

Study Record Updates

• Last Update Posted (Actual): March 25, 2025
• Last Update Submitted That Met QC Criteria: March 7, 2025
• Last Verified: March 1, 2025

More Information

Terms related to this study

• Keywords: Forensic facial reconstruction
• Additional Relevant MeSH Terms: Pathologic Processes; Disease Attributes; Facies
• Other Study ID Numbers: Forensic Facial Reconstruction; Computed FFR using CT (Other Identifier: Queen Mary University of London)