Artificial Intelligence in Echocardiography

NCT03936413 | Withdrawn FDA Device

• 1 other identifier: NYPresbyterianH
• Study Type: interventional
• Target: N/A
• Locations: 0 countries
• Sites: N/A

Brief Summary

The goal of this study is to determine whether the Bay Labs artificial intelligence (AI) system can be used by minimally trained operators to obtain diagnostic quality echocardiographic images.

Conditions

Cardiovascular Diseases

Keywords

imaging; echocardiography; artificial intelligence

Outcome Measures

Primary Outcomes (2)

• Echocardiogram image acquisition quality

  Echocardiogram image acquisition quality. View quality will be characterized by the following system: adequate (American College of Emergency Physicians (ACEP) score 5), mildly limited (ACEP score 4), moderately limited (ACEP score 3), and severely limited (ACEP score 1 and 2). This will graded for each echocardiogram view as below and for the study as a whole. * Parasternal long axis * Parasternal short axis - aortic valve level * Parasternal short axis - mid ventricle * Apical 4 chamber * Apical 2 chamber * Subcostal - 4 chamber * Subcostal - inferior vena cava

  immediately after the intervention

• Educational outcome

  The medical resident's comfort with echocardiography will be established using the following questionnaire. The answers to each question are (1) very uncomfortable, (2) somewhat uncomfortable, (3) somewhat comfortable, and (4) very comfortable. These answers will be reported separately and in aggregate. How comfortable do you feel in your knowledge of the indications for ordering an echocardiogram? How comfortable do you feel in understanding echocardiographic reports as it applies to your patients? How comfortable do you feel in obtaining routine echocardiographic views using an ultrasound machine? How comfortable do you feel in interpreting echocardiographic images after the images have already been obtained? In an emergency situation, how comfortable would you feel in performing an echocardiogram using an ultrasound machine and interpreting the images to rule out serious cardiac pathology such as severe left ventricular systolic dysfunction or a large pericardial effusion?

  1 month

Study Arms (2)

Bay Labs EchoGPS group

EXPERIMENTAL

In this arm, medical residents will use the Bay Labs EchoGPS system to perform an echocardiogram.

Device: Bay Labs EchoGPS Echcoardiogram

Native Terason group

ACTIVE COMPARATOR

In this arm, medical residents will use the native Terason machine to perform an echocardiogram.

Device: Native Terason Echocardiogram

Interventions

Bay Labs EchoGPS Echcoardiogram DEVICE

An echocardiogram will be performed in this arm using the Bay Labs EchoGPS. The Bay Labs EchoGPS system is an ultrasound system which uses the techniques of computer vision to analyze echocardiography images in real time. It then provides feedback to the user to optimize the images, and once they meet a specific level of quality it automatically records the images.

Bay Labs EchoGPS group

Native Terason Echocardiogram DEVICE

An echocardiogram will be performed in this arm using a native Terason echocardiography system. This system will not have any artificial intelligence assistance in image optimization or selection.

Native Terason group

Eligibility Criteria

• Age: 18 Years+
• Sex: all
• Healthy Volunteers: No
• Age Groups: Adult (18-64), Older Adult (65+)

You may qualify if:

• Patients must be admitted to the resident cardiology ward service under a cardiology attending (general, heart failure, or private)
• The patient must have either have undergone or be planned to undergo a formal echocardiogram within 1 day of the study echocardiogram
• The patient must consent to the study
• The patient's inpatient attending physician must give permission for the patient to be approached for consent

You may not qualify if:

• Patient refusal
• No recent or planned echocardiogram within 1 day of the study echocardiogram
• Clinical need for an emergent echocardiogram that will immediately impact clinical decision making that should instead trigger obtaining a formal echocardiogram (for example, concern for cardiac tamponade or acute myocardial infarction).

Sponsors & Collaborators

• New York Presbyterian Hospital: lead
• Caption Health, Inc.: collaborator

Related Publications (2)

• Virnig BA, Shippee ND, O'Donnell B, Zeglin J, Parashuram S. Trends in the use of echocardiography, 2007 to 2011. 2014 May 13. In: Data Points Publication Series [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2011-. Data Points #20. Available from http://www.ncbi.nlm.nih.gov/books/NBK208663/

  PMID: 24967475 BACKGROUND

• Gandhi S, Mosleh W, Shen J, Chow CM. Automation, machine learning, and artificial intelligence in echocardiography: A brave new world. Echocardiography. 2018 Sep;35(9):1402-1418. doi: 10.1111/echo.14086. Epub 2018 Jul 5.

  PMID: 29974498 BACKGROUND

MeSH Terms

Conditions

Cardiovascular Diseases

Study Officials

• Kerry A Esquitin, MD

  Columbia University

  PRINCIPAL INVESTIGATOR

Study Design

• Study Type: interventional
• Phase: not applicable
• Allocation: NON RANDOMIZED
• Masking: NONE
• Purpose: DIAGNOSTIC
• Intervention Model: SEQUENTIAL
• Sponsor Type: OTHER
• Responsible Party: PRINCIPAL INVESTIGATOR
• PI Title: Assistant Professor of Medicine

Model Details: See protocol

Study Record Dates

• First Submitted: April 16, 2019
• First Posted: May 3, 2019
• Study Start: January 13, 2020
• Primary Completion: January 13, 2020
• Study Completion: January 13, 2020
• Last Updated: December 17, 2021

Record last verified: 2021-12

Data Sharing

• IPD Sharing: Will not share