Artificial intelligence for automated evaluation of aortic measurements in 2D echocardiography: Feasibility, accuracy, and reproducibility.
Aims: This study sought to examine the feasibility, accuracy and reproducibility of a novel, fully automated 2D transthoracic echocardiography (2D TTE) parasternal long axis (PLAX) view aortic measurements quantification software compared to board‐certified cardiologists in controlled clinical setti...
| Published in: | Echocardiography Vol. 39; no. 11; pp. 1439 - 1446 |
|---|---|
| Main Authors: | , , , , , , , , , , , , , |
| Format: | diagnostic images research tables/charts Journal Article |
| Published: |
Wiley-Blackwell
Nov2022
|
| Online Access: | View this record in EBSCOhost |
| fields | @attributes: recordID: 1 pdfLink: plink: https://search.ebscohost.com/login.aspx?direct=true&db=ccm&AN=160177864&site=ehost-live header: @attributes: shortDbName: ccm uiTerm: 160177864 longDbName: CINAHL Complete uiTag: AN controlInfo: bkinfo: dissinfo: jinfo: jid: 07422822 GSE jtl: Echocardiography issn: 07422822 maglogo: Y pubinfo: dt: Nov2022 vid: 39 iid: 11 pid: 480 pub: Wiley-Blackwell place: Malden, Massachusetts artinfo: ui: 160177864 159771220 160177864 160177864 10.1111/echo.15475 160177864 ppf: 1439 ppct: 7 formats: fmt: – @attributes: type: T – @attributes: type: C – @attributes: type: P tig: atl: Artificial intelligence for automated evaluation of aortic measurements in 2D echocardiography: Feasibility, accuracy, and reproducibility. aug: au: Karužas, Arnas Balčiūnas, Justinas Fukson, Mark Verikas, Dovydas Matuliauskas, Dovydas Platūkis, Tautvydas Vaičiulienė, Dovilė Jurgaitytė, Julija Kupstytė‐Krištaponė, Nora Dirsienė, Rūta Jaruševičius, Gediminas Šakalytė, Gintarė Plisienė, Jurgita Lesauskaitė, Vaiva affil: Institute of Cardiology, Lithuanian University of Health Sciences, 15 Sukileliu street, Kaunas 50103,, Lithuania sug: subj: Artificial Intelligence Echocardiography Automation Aorta Anatomy and Histology Human Image Processing, Computer Assisted Software Descriptive Statistics Nonexperimental Studies Quantitative Studies Retrospective Design DICOM Power Analysis Repeated Measures Analysis of Variance Neural Networks (Computer) Male Female Structural Equation Modeling Reproducibility of Results Middle Age Aged Ventricular Ejection Fraction Lithuania Middle Aged: 45-64 years Aged: 65+ years Male Female ab: Aims: This study sought to examine the feasibility, accuracy and reproducibility of a novel, fully automated 2D transthoracic echocardiography (2D TTE) parasternal long axis (PLAX) view aortic measurements quantification software compared to board‐certified cardiologists in controlled clinical setting. Methods and results: Aortic Annulus (AoA), Aortic Sinus (AoS), Sinotubular Junction (STJ) and Proximal Ascending Aorta (AAo) diameter measurements were performed retrospectively on each of 58 subjects in two different ways: twice using a fully automated software (Ligence Heart version 2) and twice manually by three cardiologists (ORG) and one expert cardiologist (EC). Out of 58 studies AoA was measured in 54 (93%), AoS in 55 (95%), STJ in 55 (95%) and AAo in 54 (93%) studies. Automated measurements had a stronger correlation with EC when compared to ORG with the largest correlation difference of.1 for STJ measurements and lowest difference of.01 for AoS measurements. Automated software was in higher agreement with ground truth intervals (ORG measurements mean +‐ SEM) in three out of four measurements. Conclusion: Fully automated 2D TTE PLAX view aortic measurements using a novel AI‐based quantification software are feasible and yield results that are in close agreement with what experienced readers measure manually while providing better reproducibility. This approach may prove to have important clinical implications in the automation of the aortic root and ascending aorta assessment to improve workflow efficiency. pubtype: Academic Journal doctype: diagnostic images research tables/charts Journal Article ougenre: Article language: English refInfo: holdings: @attributes: islocal: N |
|---|