Digital Dysmorphology Project

Down Syndrome Detection From Facial Photographs Using Machine Learning Techniques

In this study, the investigators propose a novel method to detect Down syndrome using photography for facial dysmorphology, a tool called computer-aided diagnosis (CAD). After validating the method, this technology will be expanded to perform similar functions to assist in the detection of other dysmorphic syndromes.

By using photography and image analysis this automated assessment tool would have the potential to improve the diagnosis rate and allow for remote, non-invasive diagnostic evaluation for dysmorphologists in a timely manner.

Study Overview

• Status: Recruiting
• Conditions: Down Syndrome
• Intervention / Treatment: Device: photographs

Detailed Description

In this study, investigators propose a novel method to detect Down syndrome using photography for facial dysmorphology, a tool called computer-aided diagnosis (CAD) . Local texture features based on Contourlet transform and local binary pattern are investigated to represent the facial characteristics. A support vector machine classifier is then used to discriminate between normal and abnormal cases. Accuracy, precision and recall are used to evaluate the method. After validating the method, this technology will then be expanded to perform similar functions to assist in the detection of other dysmorphic syndromes.

By using photography and image analysis this automated assessment tool would have the potential to improve the diagnosis rate and allow for remote, non-invasive diagnostic evaluation for dysmorphologists in a timely manner.

• Study Type: Interventional
• Enrollment (Anticipated): 750
• Phase: Not Applicable

Contacts and Locations

• Study Contact: Name: Kevin Cleary, PhD; Phone Number: 202 476 3809; Email: kcleary@childrensnational.org
• Study Contact Backup: Name: Marius Linguraru, PhD; Phone Number: 202 476 3059; Email: MLingura@childrensnational.org

Study Locations

• United States
  • District of Columbia
    • Washington, District of Columbia, United States, 20010
      • Recruiting
      • Children's National
      • Contact: Kevin Cleary, PhD; Phone Number: 202 476 3809; Email: kcleary@childrensnational.org
      • Contact: Sara Alyamani, BS; Phone Number: 202-476-6099; Email: salyaman@childrensnational.org
      • Principal Investigator: Kevin Cleary, PhD
      • Sub-Investigator: Marius Linguraru, PhD

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: No older than 18 years (Child, Adult)
• Accepts Healthy Volunteers: No
• Genders Eligible for Study: All

Description

Inclusion Criteria:

• Pediatric subject with Down syndrome.
• Healthy pediatric siblings of a subject with Down syndrome and/or other individuals with another genetic referral to serve as a control group.
• Subject must be less than 18 years old.

Exclusion Criteria:

• Subjects 18 years or older.

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Health Services Research
• Allocation: Non-Randomized
• Interventional Model: Parallel Assignment
• Masking: None (Open Label)

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Active Comparator: Down syndrome; photographs of individuals less than 18 yo with Down syndrome	Device: photographs; computer based program to analyze photographs (computer-aided diagnosis (CAD) software)
Active Comparator: Control group; photographs of individuals less than 18 yo with a genetic referral (not Down syndrome) or a healthy sibling to a child with Down syndrome	Device: photographs; computer based program to analyze photographs (computer-aided diagnosis (CAD) software)

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Number of participants with Down syndrome accurately assessed by computer-aided detection (CADe) tool	The study will enroll and analyze photographic data from syndromic and non-syndromic cases to investigate the parameters required to achieve an accuracy of the computer-aided detection (CADe) tool for children with genetic syndromes at a level of 90% accuracy.	5 years
Number of participants with Down syndrome accurately assessed by computer-aided detection (CADe) tool	The study will enroll and analyze photographic data from syndromic and non-syndromic cases to investigate the parameters required to achieve an accuracy of the computer-aided detection (CADe) tool for children with genetic syndromes at a level of 95% accuracy.	5 years

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Number of participants with other dysmorphic syndromes accurately assessed by computer-aided detection (CADe) tool	The study will enroll and analyze photographic data from syndromic and non-syndromic cases to investigate the parameters required to achieve an accuracy of the computer-aided detection (CADe) tool for children with genetic syndromes at a level of 90% accuracy.	5 years
Number of participants with other dysmorphic syndromes accurately assessed by computer-aided detection (CADe) tool	The study will enroll and analyze photographic data from syndromic and non-syndromic cases to investigate the parameters required to achieve an accuracy of the computer-aided detection (CADe) tool for children with genetic syndromes at a level of 95% accuracy.	5 years

Collaborators and Investigators

• Sponsor: Kevin Cleary
• Collaborators: George Washington University; Children's National Research Institute; Chiang Mai University
• Investigators: Principal Investigator: Kevin Cleary, PhD, Children's National

Study record dates

Study Major Dates

• Study Start: February 1, 2013
• Primary Completion (Anticipated): December 1, 2023
• Study Completion (Anticipated): December 1, 2023

Study Registration Dates

• First Submitted: October 22, 2015
• First Submitted That Met QC Criteria: January 7, 2016
• First Posted (Estimate): January 11, 2016

Study Record Updates

• Last Update Posted (Actual): April 5, 2021
• Last Update Submitted That Met QC Criteria: April 2, 2021
• Last Verified: April 1, 2021

More Information

Terms related to this study

• Additional Relevant MeSH Terms: Pathologic Processes; Nervous System Diseases; Neurologic Manifestations; Neurobehavioral Manifestations; Disease; Congenital Abnormalities; Genetic Diseases, Inborn; Intellectual Disability; Abnormalities, Multiple; Chromosome Disorders; Syndrome; Down Syndrome
• Other Study ID Numbers: Pro00003506.