Cristin-resultat-ID: 1851457
Sist endret: 24. november 2020, 10:30
Resultat
Doktorgradsavhandling
2020

Automatic measurements of mitral annular motion indices

Bidragsytere:
  • Jahn Frederik Grue

Utgiver/serie

Utgiver

NTNU

Serie

Doktoravhandlinger ved NTNU
ISSN 1503-8181
NVI-nivå 0

Om resultatet

Doktorgradsavhandling
Publiseringsår: 2020
Hefte: 4
Antall sider: 137
ISBN: 978-82-326-3912-0

Klassifisering

Fagfelt (NPI)

Fagfelt: Hjerte, kar og luftveier
- Fagområde: Medisin og helsefag

Beskrivelse Beskrivelse

Tittel

Automatic measurements of mitral annular motion indices

Sammendrag

Echocardiography (ultrasound examination of the heart) is essential for evaluating cardiac function. As a result of hand-held ultrasound devices, ultrasound is now available outside of hospitals, but in order to fully profit from the potential benefits of the widespread availability, the imaging process should be simplified. Interpretation of ultrasound images is challenging. A solution can be to introduce automatic image analyses that let the examiner rely on quantitative information instead of subjective (visual) assessment. Left ventricular longitudinal function is reduced in a wide range of cardiac diseases, and this function can be quantified by measuring mitral annular motion indices. These indices are robust against low image quality and are suitable for being measured automatically. In this project, an algorithm for fully automatic echocardiographic measurements of mitral annular motion indices was developed and tested. The algorithm operated on recordings from high-end scanners, but it can be implemented in future hand-held ultrasound devices. The aims were to evaluate the quality of automatic measurements in cardiac patients and healthy subjects when the measurements were obtained from recordings acquired by operators with high and low experience level in ultrasound, to create normal reference values, and to evaluate the automatic measurements’ potential for detection of left ventricular dysfunction. The work consists of four studies. Study 1 included both patients referred to echocardiography and healthy subjects. Mitral annular motion indices were measured automatically in ultrasound recordings acquired by highly competent echocardiographers. There was a close agreement between automatic measurements and manual reference measurements. Patients referred to echocardiography participated in Study 2. Medical students acquired ultrasound recordings from the patients and the recordings were analyzed automatically for mitral annular motion indices. The automatic measurements were not sufficiently similar to manual reference measurements from the echocardiographic examinations, probably due to low image quality in many of the student recordings. This underlines that thorough training in image acquisition is necessary for valid automatic analyses. In Study 3, echocardiographic examinations from a large, healthy population was utilized to create ageand gender-specific normal ranges for automatic measurements of mitral annular motion indices. These ranges for normalcy enable more precise interpretation of left ventricular function. Detection of abnormal left ventricular function using the automatic measurements was evaluated in Study 4. Patients referred for echocardiography were included. Classification of left ventricular function based on the clinical echocardiogram was used as the reference. The study indicated that automatically measured systolic mitral annular motion indices could reveal left ventricular dysfunction in a heterogeneous group of cardiac patients.

Bidragsytere

Jahn Frederik Noraas Grue

Bidragsyterens navn vises på dette resultatet som Jahn Frederik Grue
  • Tilknyttet:
    Forfatter
    ved Institutt for sirkulasjon og bildediagnostikk ved Norges teknisk-naturvitenskapelige universitet

Bjørn Olav Haugen

  • Tilknyttet:
    Veileder

Håvard Dalen

  • Tilknyttet:
    Veileder
    ved Institutt for sirkulasjon og bildediagnostikk ved Norges teknisk-naturvitenskapelige universitet

Hans Torp

  • Tilknyttet:
    Veileder
    ved Institutt for sirkulasjon og bildediagnostikk ved Norges teknisk-naturvitenskapelige universitet
1 - 4 av 4