Development of an AI Solution to Evaluate Clinical Activity of TAO Using Periocular Images Taken by Smartphones in Patients With Thyroid Dysfunction

Development of an Artificial Intelligence Solution to Evaluate Clinical Activity of Thyroid Associated Orbitopathy Using Periocular Selfie Images Taken by Smartphones in Patients With Thyroid Dysfunction

This study was conducted to develop an AI solution to evaluate clinical activity score (CAS) of thyroid associated orbitopathy (TAO) using periocular selfie images taken by smartphones in patients with thyroid dysfunction.

An AI solution to evaluate CAS in TAO was developed in previous retrospective study using periorbital digital images taken by DSLR camera in the studio. This AI solution will be adjusted to show its best performance with periocular selfie images taken by smartphones.

Study Overview

• Status: Completed
• Conditions: Thyroid Eye Disease
• Intervention / Treatment: Other: taking periorbital selfie images using smartphones

• Study Type: Observational
• Enrollment (Actual): 100

Contacts and Locations

Study Locations

• Korea, Republic of
  • Gyeonggi-do
    • Seongnam-si, Gyeonggi-do, Korea, Republic of, 13620
      • Seoul National University Bundang Hospital

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 19 years and older (Adult, Older Adult)
• Accepts Healthy Volunteers: No
• Genders Eligible for Study: All

• Sampling Method: Non-Probability Sample

Study Population

This study is conducted to develop AI solution using smartphone selfie images. Therefore, the number of subjects for this study was calculated as 200, the maximum number of patients with thyroid eye disease that can be prospectively recruited during the study period.

Description

Inclusion Criteria:

• Patients with thyroid eye disease

Exclusion Criteria:

• Patients who have prior facial anomaly or surgery
• Patients who cannot use smartphone camera app

Study Plan

How is the study designed?

Number of groups / cohorts

1

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
thyroid eye disease group; Patients who were diagnosed as thyroid eye disease. Periorbital images are taken by DSLR camera in the studio. Periorbital selfie images are taken by the patients with 6 smartphones including various camera modules. CAS of TAO is evaluated by 3 ophthalmologists based on the periorbital digital images taken by DSLR camera and symptoms.	Other: taking periorbital selfie images using smartphones; taking periorbital selfie images using smartphones

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Accuracy for clinical activity score (CAS) of thyroid eye disease (TED)	Accuracy of AI solution using smartphone selfie images for CAS of TED.	baseline (this study has no intervention)
Diagnostic accuracy for active thyroid eye disease	Accuracy of AI solution using smartphone selfie images for active thyroid eye disease (TED) which is defined as 3 or more of clinical activity score	baseline (this study has no intervention)

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Accuracy for redness of eyelids	Accuracy of AI solution using smartphone selfie images for redness of eyelids	baseline (this study has no intervention)
Accuracy for redness of conjunctiva	Accuracy of AI solution using smartphone selfie images for redness of conjunctiva	baseline (this study has no intervention)
Accuracy for swelling of eyelids	Accuracy of AI solution using smartphone selfie images for swelling of eyelids	baseline (this study has no intervention)
Accuracy for swelling of conjunctiva	Accuracy of AI solution using smartphone selfie images for swelling of conjunctiva	baseline (this study has no intervention)
Accuracy for swelling of caruncle	Accuracy of AI solution using smartphone selfie images for swelling of caruncle	baseline (this study has no intervention)

Collaborators and Investigators

• Sponsor: Seoul National University Bundang Hospital
• Collaborators: THYROSCOPE INC.
• Investigators: Principal Investigator: Namju Kim, MD, PhD., Seoul National University Bundang Hospital

Study record dates

Study Major Dates

• Study Start (Actual): December 22, 2021
• Primary Completion (Actual): July 18, 2022
• Study Completion (Actual): July 25, 2022

Study Registration Dates

• First Submitted: April 17, 2022
• First Submitted That Met QC Criteria: February 15, 2023
• First Posted (Estimate): February 16, 2023

Study Record Updates

• Last Update Posted (Estimate): February 16, 2023
• Last Update Submitted That Met QC Criteria: February 15, 2023
• Last Verified: February 1, 2023

More Information

Terms related to this study

• Keywords: machine learning; smartphone; artificial intelligence; thyroid eye disease
• Additional Relevant MeSH Terms: Immune System Diseases; Autoimmune Diseases; Endocrine System Diseases; Genetic Diseases, Inborn; Eye Diseases, Hereditary; Graves Disease; Exophthalmos; Orbital Diseases; Goiter; Hyperthyroidism; Eye Diseases; Graves Ophthalmopathy; Thyroid Diseases
• Other Study ID Numbers: B-2110-715-304

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO
• IPD Plan Description: IPD will not be shared.

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No