Quantification of aortic valve stenosis based on a time-frequency analysis of doppler ultrasound signals

J. H. Nagel, S. P. Reddy, B. E. Hurwitz, N. Schneiderman, R. Prashad, Arthur Agatston

Research output: Contribution to journalArticlepeer-review


Echocardiographic Determination of the transvalvular pressure gradient has become the method of choice for non-invasive evaluation of aortic valvular stenosis. Quantification of aortic stenosis based on the pressure gradient, however, has been questioned as to its reliability, especially in routine applications, in the presence of substantial beat-to-beat variations in stroke volume as caused by arrhythmia, and in patients with some degree of accompanying left ventricular failure(LVF). Thus, in order to back up the results obtained with the gradient technique, i.e. to eliminate possible errors due to its shortcomings, and to determine the degree of stenosis with higher reliability and accuracy, the valve area needs to be measured directly, by transesophageal US or to be calculated by the less accurate continuity equation method. A procedure for the morpholgical evaluation of the continuous wave(CW) doppler velocity spectra based on a time-frequency analysis has been developed which provides a reliable method of quantifying the degree of stenosis that is not affected by the shortcomings of pressure gradient evaluation and eliminates the need of valve area determination. The time-frequency spectrum of the US signal for a single heart beat reveals two distinct peaks. These peaks move closer together with increasing degree of stenosis, showing some overlapping in cases of severe stenosis. The ratio of the distance between the two peaks to the left ventricular ejection time, called shape factor(SF), represents an unequivocal, reliable, and accurate signature for the shape of the CW velocity spectra which is independent of the pressure gradient, even in cases of LVF and atrial fibrillation. Also, SF can be determined automatically without operator involvement, thus eliminating problems of less reliable routine measurements as inherent to the gradient and valve area techniques. The shape factor strongly correlates with the degree of stenosis. In clinical tests, SF has not only been able to distinguish among groups of patients which had been classified by an expert as normal, mild and severely stenotic, SF has also been proven to establish an objective, reliable, accurate, and continuous scale for the quantification of aortic valve stenosis. Analysis of data obtained from 31 patients showed a correlation of r2=0.88 between SF and valve area as determined by transesophageal US and the continuity equation.

Original languageEnglish (US)
Pages (from-to)1-7
Number of pages7
JournalBiomedical Engineering - Applications, Basis and Communications
Issue number2
StatePublished - Aug 19 1997
Externally publishedYes


  • Aortic Valve
  • Doppler Ultrasound
  • Stenosis
  • Time-Frequency Analysis
  • Valvular Aortic Steosis

ASJC Scopus subject areas

  • Biophysics
  • Bioengineering


Dive into the research topics of 'Quantification of aortic valve stenosis based on a time-frequency analysis of doppler ultrasound signals'. Together they form a unique fingerprint.

Cite this