CARDIOVASCULAR SOUND 



683 



a pressure transducer between a drumstick and a 

 drum head. It would sense the forces which create the 

 sound, but would not record the sound generated. 



So one can hardly say that the sonic catheter 

 senses cardiac sounds at their origin; but rather a 

 complex of hydrodynamic flow pressures associated 

 in various ways with their generation. This is not to 

 minimize the importance of such observations in de- 

 scribing anomalies, but rather to avoid the irrational 

 attempt to find correspondences of form between 

 "sound" patterns recorded externally and internally. 

 Because these vortical patterns are highly nonlinear, 

 and spatially complex, any sound waves arising from 

 their vicinity would have to be evoked rather than 

 driven by them, much as with the blow on a drum. 



The dynamic range of cardiova.scular sound is con- 

 siderable. In the case of the normal heart the range 

 is as much as 80 db between the intense components 

 at 10 cps and the extremely faint components with 

 intensities of the order of the threshold sensitivity at 

 the level of 1000 cps. In pathologic situations the 

 range may be considerably greater. 



The decibel (db) is a unit u.sed to express power 

 ratios such that 



db = 10 1o:t, 



Pi 



A positive sign or no sign preceding the db value 

 indicates Pi is greater than P>. The usefulness of db 

 measurements comes from the ease with which large 

 ratios can be expressed. Thus, 40 db represents a 

 power ratio of 10,000. In electrical circuits the power 

 is proportional to the square of the voltage or the 

 current, and in acoustical measurements the intensity 

 of .sound is proportional to the square of pressure. 

 Because of this .second power relationship 



db = 20 \os — or db = 20 loe — 

 I '2 P' 



Care must be exercised both in using db and inter- 

 preting how others have used it (i). Although origi- 

 nally it was applied only to constant impedance 

 circuits the usage has been extended to ratio measure- 

 ments of all kinds. However, the 10 log factor is 

 retained for all power measurements and 20 log 

 factor for units such as voltage, current, or pressure. 

 For instance, a voltage amplifier may have a stated 

 gain of 40 db. That is a voltage gain of 100. The 

 power gain may not be stated or may not be of interest 

 and, in fact, will probably not be 40 db. 



Since the db unit is a ratio, it cannot be used to 

 indicate an absolute value. However, in practice, 

 there are several situations in which a reference level 

 is assumed so that the db unit is in effect used to 

 indicate an ab.solute value. For instance, a high 

 impedance microphone (6) may have a stated sensi- 

 tivity of —60 db. This means that its voltage output 

 is —60 db or i/iooo relative to one volt per dyne 

 per square centimeter of incident pressure. The refer- 

 ence level of I volt per dyne per square centimeter 

 may or may not be stated. 



Similarly, sound pressure levels are often expressed 

 in db with an implied reference level of 0.0002 dynes 

 per cm-, the accepted threshold of hearing at 1000 

 cycles per sec. 



Low impedance microphones ( i ) may have a listed 

 sensitivity expressed in dbm (output power) re 10 

 dynes per cm^ (input pressure). In this case the power 

 is referred to i milliwatt, hence dbm (decibel-milli- 

 watt). To find the output voltage, the impedance 

 must be stated also. Transformers are often rated in 

 dbm. 



It can be seen that the use of the db unit is wide- 

 spread and can be useful. However, care must be 

 exercised in stating the reference for the ratio measure- 

 ment. Before making a plot in db, one should consider 

 whether or not a plot in absolute units would be more 

 helpful to the reader even though it may seem less 

 sophisticated. 



Due to the intensity relation shown in figure 3, the 

 frequency range of displayed cardiovascular sound 

 from the normal heart rarely extends higher than 

 1000 cps. In pathologic conditions resulting in much 

 accentuated heart sounds and very loud murmurs the 

 peak frequency range may exceed 1000 cps and even 

 reach 2000 cps. Although only a small proportion of 

 the total energy is distributed at frequencies above 

 1000 cps, one should not conclude it is without 

 physiologic or clinical importance. 



Analysis of results, such as those of Mannheimer, 

 indicate that cardiovascular sound has a natural 

 decrement of about 12 db per octave or 40 db per 

 decade. [This is the physiologic decrement referred 

 to by Zalter et al. (19). It means, in the opinion of the 

 writers, that an additional physical decrement referred 

 to by Zalter et al. is not operative in the recording of 

 cardiovascular sound.] Considering the total spectrum 

 of precordial x-ibration in a unified manner, a decre- 

 ment of 40 db per decade represents a dynamic 

 range of 120 db in the band from i to 1000 cps, and 

 80 db in the band from 10 to 1000 cps. 



