Spectral phonocardiographic studies in congenital heart disease.

The principles and technique of spectral phonocardiography have been described in detail in earlier publications from this laboratory (McKusick et al., 1954). The method provides detailed frequency-time-intensity plots of cardiovascular sound. It resembles in some respects the manner of operation of the auditory mechanism on these sounds. In essence the spectral phonocardiogram is made up of a large number of horizontal lines, each representing the output of an individual pass-band filter. In other words, in making the spectral phonocardiogram a selected segment of sound (in electrical form) is in effect put through a battery of a very large number of individual electrical filters each of which is tuned to a specific and different, although overlapping, pass-band. Spectral phonocardiography has the following advantages over conventional oscillographic phonocardiography (McKusick et al., 1955): (1) resolution in the time dimension is improved; (2) the wide intensity range of cardiovascular sound is better encompassed with simultaneous display of faint and loud components in their true proportions; (3) quality (timbre) is represented accurately; and (4) vibrations of non-cardiovascular origin are more easily identified. Spectral phonocardiography can do, or at least can be made to do, all the ear can. It can exceed the performance of the ear because (1) resolution in the time dimension is better, (2) it is not wed to one particular frequency response curve, (3) it is not handicapped by psycho-acoustic phenomena such as masking, and (4) permanent, quantifiable records are produced.