Feasibility of AI-based Heart Function Prediction Model Using CXR

Chest X-ray for Clinical Evaluation Clinical Trial

— AI-CXR  

Official title:

Feasibility of Artificial Intelligence-based Heart Function Prediction Model Using Chest Radiography

Clinical Trial Details — Status: Completed

Administrative data

• NCT number: NCT04996381
• Other study ID #: YonseiU
• Status: Completed
• Start date: March 1, 2022
• Est. completion date: September 1, 2022

Study information

• Verified date: September 2022
• Source: Yonsei University
• Contact: n/a
• Is FDA regulated: No
• Study type: Observational

Clinical Trial Summary

The investigators will develop an artificial intelligence model to predict left ventricular ejection fraction using chest radiographic images and transthoracic echocardiography data.

Description:

Echocardiography should be considered at an early stage in patients who have first developed heart failure or who do not have information about heart function, but the examination may be delayed due to lack of time and manpower in the actual medical field.

Primary Objective: Use chest radiographs to predict the left ventricular ejection fraction

Recruitment information / eligibility

• Status: Completed
• Enrollment: 505
• Est. completion date: September 1, 2022
• Est. primary completion date: June 30, 2022
• Accepts healthy volunteers: No
• Gender: All
• Age group: 20 Years to 90 Years

Eligibility

Inclusion Criteria:

• Adults who are 20 years and older

• Patient who visited the emergency room or outpatient clinic due to dyspnea and chest pain

Exclusion Criteria:

• Patient refusal

• Uncertain radiographs or transthoracic echocardiography

• Uncertain tests results

Related Conditions & MeSH terms

• Chest X-ray for Clinical Evaluation

Intervention

Diagnostic Test:
• Scanning Chest X-rays and performing AI algorithms on images
Chest X-Rays; AI CNNs; Results

Locations

Country	Name	City	State
Korea, Republic of	Yongin Severance Hospital	Yongin	Giheung-gu

Sponsors (1)

Lead Sponsor	Collaborator
Yonsei University

Country where clinical trial is conducted

Korea, Republic of, 

Outcome

Type	Measure	Description	Time frame	Safety issue
Primary	Left Ventricular Ejection Fraction < 40%	Evaluate the performance of chest X-ray based artificial intelligence algorithms to identify individuals with reduced ejection fraction (<40%)	Within two weeks of chest X-ray