Premium
Image Quality Influences the Assessment of Left Ventricular Function
Author(s) -
Grossgasteiger Manuel,
Hien Maximilian D.,
Graser Bastian,
Rauch Helmut,
Motsch Johann,
Gondan Matthias,
Rosendal Christian
Publication year - 2014
Publication title -
journal of ultrasound in medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.574
H-Index - 91
eISSN - 1550-9613
pISSN - 0278-4297
DOI - 10.7863/ultra.33.2.297
Subject(s) - medicine , ventricular function , quality (philosophy) , function (biology) , cardiology , image quality , artificial intelligence , image (mathematics) , philosophy , epistemology , evolutionary biology , biology , computer science
Objectives Transesophageal echocardiography has become a standard tool for eval uating left ventricular function during cardiac surgery. However, the image quality varies widely between patients and examinations. The aim of this study was to investigate the influence of the image quality on 5 commonly used 2‐dimensional methods. Methods Transesophageal real‐time 3‐dimensional echocardiography (3DE) served as a reference. Left ventricular function was evaluated in 63 patients with sufficient real‐time 3DE image quality. The image quality was rated using the ratio of the visualized border divided by the total endocardial border. These ratings were used to generate groups of poor (0%–40%), fair (41%–70%), and good (71%–100%) image quality. The ejection fraction (EF), end‐diastolic volume, and end‐systolic volume were analyzed by the Simpson method of disks (biplane and monoplane), eyeball method, Teichholz method, and speckle‐tracking method. Furthermore, the fractional area change was determined. Each correlation with real‐time 3DE was evaluated. Results Correlations of the EF and volumes, respectively, as determined by long‐axis view methods increased with improving image quality: the Simpson biplane method was found to be the most accurate method, with good image quality for the EF ( r = 0.946) and volumes (end‐diastolic volume, r = 0.962; end‐systolic volume, r = 0.989). Correlations of the EF and fractional area change by short‐axis view methods decreased with improving image quality, with the Teichholz EF found to be most accurate with poor ( r = 0.928) in contrast to good ( r = 0.699) image quality. Conclusions With good image quality, the Simpson biplane method is the most accurate 2‐dimensional method for assessing the left ventricular EF. Short‐axis view methods, especially the Teichholz method yield better correlations with poor image quality. The eyeball method was unaffected by image quality.