CARDIOVASCULAR SOUND 



693 



significance, and because the weight of the micro- 

 phone has a "ioading" effect. Some (2) have used a 

 suction cup to hold the microphone to the surface 

 from wliicli recording was to be done. 



Intracardiac phonocardiography had its crude begin- 

 nings with the intracardiac pressure recordings of 

 MacLeod & Cohn (10), who found high-frequency 

 vibrations synchronous to the heart .sounds super- 

 imposed on the pressure trace. Recording of the rapid 

 pressure transients within the heart has been per- 

 formed by means of a barium titanate' (17) or minia- 

 ture electromagnetic" (7, 16) transducer on the tip of 

 a catheter, by a condenser microphone with one pole 

 at the tip of the catheter and the other the body 

 itself (18), or a transducer located on the external 

 end of the catheter and communication with the 

 interior via the blood filling the catheter. If one uses 

 an internal transducer, a double lumen catheter 

 permits one to combine intracardiac phonocardiog- 

 raphy with pressure recording, sampling of blood, 

 and administration of indicator dilution agents. 



Information obtained with the transducer at the 

 external end of a catheter must be interpreted with 

 exceeding caution. Intracardiac transducers of the 

 barium titanate type are disturbingly sensitive to 

 bending of the catheter (4), to impacts at a distance 

 from the tip, and even to minor temperature changes. 

 On the other hand, they are relatively insensitive to 

 localized pressure changes at the tip. 



The analysis of heart sounds by digital computer 

 techniques appears to have promise. For example, in 

 determining the ""normar" spectral phonocardiogram, 

 difficulties result from the wealth of information con- 

 tained in the recording. A possible solution for this 

 "embarrassment of riches" is the translation of the 

 information into the digital language of the computer 

 and the use of machine methods for determining the 

 biological variability of the normal. 



Other special techniques such as esophageal 

 phonocardiography, fetal phonocardiography, and 

 others require no special consideration here. 



CALIBR.^'iTION OF INTENSITY 



Intensity calibration in phonocardiography pre- 

 sents many problems. There are variables in the 



' The barium titanate microphone and preamphfier used by 

 Lewis and his colleagues (17) are available from the American 

 Electronics Laboratories, Inc., Philadelphia, Pa. The .Allard- 

 Laurens micromanometer system used by Soulie and his col- 

 leagues (16) is available from Telco Inc., Gentilly, .Seine, 

 France. 



sound-transmitting properties of the chest among 

 subjects and in the same subject under different 

 circumstances. There can be \ariabilit\- in the per- 

 formance of the microphone which is often difficult 

 to calibrate in the first place. As mentioned earlier, 

 even the pressure with which the microphone is 

 applied to the skin surface introduces variability. As 

 pointed out, in the analyses of Rappaport & Sprague 

 (14), the influence of different chest pieces, with and 

 without diaphragm, is considerable. The electronic 

 amplification requires calibration. Changes in in- 

 tensitv must be considered in relation to frequency. 

 Sounds of quite different frequency composition may 

 be of identical over-all intensity. 



The use of calibrating signal in the form of a 

 vibration generator strapped elsewhere on the chest 

 or even placed in the esophagus has insufficient com- 

 parability to the real situation to be reliable. Others 

 have calibrated from the microphone by imposing a 

 calibrating signal just proximal to the microphone. 



The idealized system (for oscillographic phono- 

 cardiography), shown in figure 3, permits semi- 

 quantitative calibration, as was indicated previously. 

 We shall probably have to be satisfied with rather 

 qualitative and, at the best, semiquantitative state- 

 ments about intensity. The use of a roughly calibrated 

 gray scale with spectral plionocardiograms is a be- 

 ginning. For clinical use, precise quantitation is 

 probably not very important. 



Careful specification of the characteristics of the 

 phonocardiographic system used in any study is 

 essential and should be clearly stated in reports. 

 [Recommendations for methods in indicating the 

 frequency characteristics of systems have been made 

 (5).] However, it can be seriously questioned whether 

 the time is ripe for standardization of technique. 

 Admittedly, it will be a convenience if many labora- 

 tories can use a few well-specified systems and in 

 their reports merely refer to the published charac- 

 teristics of the particular system used. For example, 

 much phonocardiography in this country has been 

 with the Sanborn Twin-Beam (and its predecessor 

 the Sanborn Stethocardiette), the characteristics of 

 which were rather completely determined by Rap- 

 paport. 



In summary, a five-filter system for oscillographic 

 phonocardiography, such as that schematized in 

 figure 3, can provide most of the information in- 

 corporated in the spectral phonocardiogram. It is a 

 less elegant approach than the spectral phonocardio- 

 gram which provides the full picture in a single 

 display, together with a wealth of detail which is 

 necessary for a precise visual impression of timbre. 



