Semi-automated Segmentation and Quantification of Mitral Annulus and Leaflets from Transesophageal 3-D Echocardiographic Images

Miguel Sotaquira, Mauro Pepi, Laura Fusini, Francesco Maffessanti, Roberto M. Lang, Enrico G. Caiani

Research output: Contribution to journalArticlepeer-review

Abstract

Quantification of three-dimensional (3-D) morphology of the mitral valve (MV) using real-time 3-D transesophageal echocardiography (RT3-D TEE) has proved to be a valuable tool for the assessment of MV pathologies, but of limited use in clinical practice because it relies on user-intensive approaches. This study presents a new algorithm for the segmentation and morphologic quantification of the mitral annulus (MA) and mitral leaflets (ML) in closed valve configuration from RT3-D TEE volumes. Following initialization, the MA and the ML and the coaptation line (CL) are automatically obtained in 3-D. Validation with manual tracings was performed on 33 patients, resulting in segmentation errors in the order of 0.7mm and 0.6mm for the MA and ML segmentation, in addition to good intra- and inter-observer reproducibility (coefficients of variation below 12% and 15%, respectively). The ability of the algorithm to assess different MV pathologies as well as repaired valves with implanted annular rings was also explored. The reported performance of the proposed fast, semi-automated MA and ML quantification makes it promising for future applications in clinical settings such as the operating room, where obtaining results in short time is important.

Original languageEnglish
Pages (from-to)251-267
Number of pages17
JournalUltrasound in Medicine and Biology
Volume41
Issue number1
DOIs
Publication statusPublished - Jan 1 2015

Keywords

  • Block-matching
  • Echocardiography
  • Graph-based segmentation
  • Mitral annulus
  • Mitral leaflets
  • Mitral valve
  • Mitral valve quantification

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Radiological and Ultrasound Technology
  • Biophysics

Fingerprint

Dive into the research topics of 'Semi-automated Segmentation and Quantification of Mitral Annulus and Leaflets from Transesophageal 3-D Echocardiographic Images'. Together they form a unique fingerprint.

Cite this