Assessment and validation of a novel fast fully automated artificial intelligence left ventricular ejection fraction quantification software

• DOI: 10.1111/echo.15318
• Publication Date: 2022-02-18T02:59:07Z
• Authors: Rajeev Samtani; Solomon Bienstock; Ashton C. Lai; Steve Liao; Usman Baber; Lori Croft; Eric Stern; Frans Beerkens; Peter Ting; Martin E. Goldman
• Keywords: Cardiac magnetic resonance; Bland–Altman plot

Quantification of left ventricular ejection fraction (LVEF) by transthoracic echocardiography (TTE) is operator-dependent, time-consuming, and error-prone. LVivoEF DIA a new artificial intelligence (AI) software, which displays the tracking endocardial borders rapidly quantifies LVEF. We sought to assess accuracy compared cardiac magnetic resonance imaging (cMRI) as reference standard compare standard-of-care physician-measured LVEF (MD-EF) including studies with ultrasound enhancing agents (UEAs).In 273 consecutive patients, we MD-EF AI-derived cMRI. was obtained from non-UEA four-chamber view without manual correction. Thirty-one patients were excluded: 25 had interval interventions or incomplete TTE cMRI six uninterpretable apical views.In 242 subjects, correlation between AI r = .890, similar .891 (p 0.48). Of 126 performed UEAs, using unenhanced .89, .90. In 116 studies, .87, .84. From Bland-Altman analysis, underreported bias 3.63 ± 7.40% EF points while .33 7.52% 0.80).Compared cMRI, can accurately quantify Thus, has ability improve expedite quantification.