Author + information
- Received July 19, 1982
- Revision received November 23, 1982
- Accepted November 30, 1982
- Published online May 1, 1983.
- Gary R. Johnson, MD,
- Robert J. Adolph, MD, FACC* and
- Donald J. Campbell, EE
- ↵*Address for reprints: Robert J. Adolph, MD, Division of Cardiology, University of Cincinnati College of Medicine, Mail Location 542, Cincinnati, Ohio 45267.
This study of 23 patients was designed to test the hypothesis that the severity of aortic stenosis can be estimated by application of the principle that higher murmur frequencies are generated as severity of the stenosis increases. The frequency content of the systolic murmur of aortic stenosis was determined using fast Fourier transform spectral analysis and correlated blindly with the transvalvular peak systolic pressure gradient. The 23 patients averaged 52 years of age (range 29 to 70). The systolic pressure gradient ranged between 10 and 140 mm Hg. After cardiac catheterization, the electrocardiogram and sound vibrations from the aortic area and cardiac apex were recorded on tape. The R wave initiated analog to digital conversion for the duration of the murmur. Frequency spectra of 10 murmurs were computed to obtain an average spectrum, which was normalized to minimize coupling and transmission variability.
The plot of murmur frequency versus magnitude from 25 to 75 Hz (constant area) and that from 75 to 150 Hz (predictive area) were computer-integrated. The integrated areas were normalized for comparison of patients by calculating the ratio of predictive area to constant area. This ratio recorded at the aortic area increased linearly as systolic pressure gradient increased (r = 0.90, p < 0.001); at the cardiac apex this ratio did not correlate with the systolic pressure gradient. The predictive/constant area ratio at the aortic area correlated less well with calculated valve area and the degree of calcification, and was independent of the degree of regurgitation. The severity of aortic stenosis may be underestimated by the predictive/constant area ratio in patients with a low cardiac output, in whom the systolic pressure gradient would also be reduced. Nevertheless, this noninvasive technique may prove useful in predicting the need for cardiac catheterization and in longitudinal follow-up of a majority of patients with aortic stenosis.
- Received July 19, 1982.
- Revision received November 23, 1982.
- Accepted November 30, 1982.
- American College of Cardiology Foundation