Accuracy of mitral valve area measurements using transthoracic rapid freehand 3-dimensional scanning: comparison with noninvasive and invasive methods

J Am Soc Echocardiogr. 2003 Dec;16(12):1292-300. doi: 10.1067/j.echo.2003.07.005.


Objective: The feasibility and accuracy of direct transthoracic 3-dimensional (3D) mitral valve area (MVA) measurements obtained using freehand scanning was investigated in patients with mitral stenosis.

Methods: A total of 30 patients (26 women, 4 men; aged 55 +/- 13 years) underwent a 2-dimensional (2D) and Doppler study 1 hour before percutaneous balloon mitral valvuloplasty. Transthoracic freehand data were acquired using a magnetic receiver attached to a broadband transducer, gated to electrocardiography and respiration. Volumetric MVA measurements from the left ventricle and left atrium were obtained and compared with MVA measurements derived from 2D planimetry, pressure half-time, and proximal isovelocity surface area. Invasive Gorlin MVA measurements were the gold standard for comparison.

Results: In all, 29 patients (97%) had 3D data allowing MVA measurements. Direct 3D measurements from the left ventricle had the least bias (0.06 +/- 0.19 cm(2)) and tightest limits of agreement (-0.44 to 0.32) compared with left atrium measurements (0.17 +/- 0.25 cm(2) and -0.67 to 0.33, respectively). The proximal isovelocity surface area method (bias: 0.09 +/- 0.34 cm(2)) was the most accurate of all 2D methods followed by pressure half-time (0.17 +/- 0.36 cm(2)) and planimetry (0.21 +/- 0.29 cm(2)).

Conclusion: Direct 3D MVA measurements from the left ventricle using transthoracic freehand scanning are more accurate than traditional 2D methods.

Publication types

  • Comparative Study

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Echocardiography, Doppler / methods*
  • Echocardiography, Three-Dimensional / methods*
  • Female
  • Humans
  • Image Processing, Computer-Assisted
  • Male
  • Middle Aged
  • Mitral Valve / diagnostic imaging*
  • Mitral Valve Stenosis / diagnostic imaging*
  • Observer Variation