Chapter 2. Phonocardiography

Cardiac auscultation is an important and basic component of a physical examination in clinical practice. For the experienced listener, auscultation provides quick and reliable diagnostic information about the state of the heart. Phonocardiography systems allow the listener to record these findings. Spectral phonocardiographic studies, introduced as early as 1955 by McKusick et al,1 accurately characterize the quality of heart sounds and cardiac murmurs by creating a graphic record. Recording heart sounds and phonocardiograms enables building heart sound libraries and teaching material.2-4 Dr. Proctor Harvey's audio tapes are classics in adult cardiology.

The significant improvements in personal computers over the past few decades have made it possible to design new highly capable phono-spectrocardiographic devices with digital signal analysis. Published studies have demonstrated the clinical validity of this method of recording and analyzing cardiac sounds.5-9 Digital signal processing allows the combination of phonocardiography with other imaging modalities.

Clinical Auscultation

Heart sounds provide information about both normal and pathologic physical events in the heart. The movement of blood and the movement and vibration of the heart structures produce audible sounds and murmurs. The causes of vibratory movement in the cardiovascular system can be categorized into two general mechanisms: (1) rapid acceleration and deceleration of blood flow or (2) continuous turbulence in blood flow. The sudden sounds caused by acceleration and deceleration of blood flow are called heart sounds, and the sounds produced by turbulent blood flow are labeled murmurs.

Conventional Phonocardiography

Phonocardiography is a diagnostic technique that graphically records cardiac acoustic phenomena.10 Visualization of heart sounds may help in understanding cardiac events (Fig. 2–1). A visual record may also help an inexperienced listener recognize and classify auscultatory findings. The essential elements of the phonocardiographic system are a transducer (microphone), an amplifier, a filter, and recording, analyzing, and transcription systems. In the past, all of the parts, including those needed for recording and transcription, were analog.

Digital Phonocardiography

Today, recording and analysis are done digitally using personal computers and specialized software or by using custom electronic devices specially designed for this purpose. In addition, sophisticated analysis software, feature extraction, and classification algorithms (such as neural networks) may be used for further interpretation of sounds.11,12

A visual presentation provides an opportunity to study the timing, frequency, and intensity of different heart sound and murmur components. Spectral phonocardiography combined with traditional phonocardiographic tracing simultaneously depicts all of these features. Digital signal processing allows accurate fine-tuning of the visual representation in terms of various frequencies, intensities, and timing.

Heart Sound Analysis in Children

Over 90% of heart murmurs in children are physiologic and benign. Moreover, 75% of these murmurs are innocent vibratory murmurs.13 Visual analysis of the phono-spectrocardiographic record enables differentiation between pathologic and physiologic pediatric murmurs. Simple quantitative criteria applied to the phono-spectrocardiographic record may significantly improve the diagnostic accuracy of any screener to the level of a subspecialist.9 Advanced signal processing and pattern recognition tools have yielded even better diagnostic accuracy.14 Artificial neural network–based screening has been reported to have higher sensitivity and specificity than other signal processing methods.8,12

Heart Sound Analysis in Adults

Phono-spectrocardiographic signal analysis is a promising clinical application in the assessment of the severity of aortic or carotid arterial stenosis.2,5,6 Furthermore, as in children, it can be used to differentiate between pathologic and innocent murmurs. Traditional phonocardiography combined with spectral digital analysis, also known as acoustic cardiography, has been used to recognize and quantify heart sounds S3 and S4.15-18 Changes in the recorded magnitude of S3 and/or S4 can perhaps be used to determine whether the patient's hemodynamic status is improving or worsening (Fig. 2–2).

Examples of Heart Sound Analysis

The following examples of heart sound analysis are from pediatric patients, and they demonstrate the validity of modern phono-spectrocardiographic imaging techniques. The software used includes corresponding audio recordings and marker measures (Fig. 2–3).

In these phono-spectrograms (Figs. 2–4, 2–5, 2–6, 2–7, 2–8, 2–9, 2–10, 2–11), timing is read from the time scale (at the bottom of the figures) by visually examining both the phonocardiogram and the spectrogram and then deciding the duration of a sound event or murmur and dragging the marker over the selected time segment. The intensities of S1, S2, and murmurs are read from the scale (to the right in the figures, percentages of the recording capacity of the stethoscope) by positioning the marker over the phonocardiogram. The loudness of a murmur is estimated from the phonocardiogram by comparing the maximum amplitude of the murmur to the average value of the maximum amplitude of the first and second heart sounds. Finally, the frequencies (in hertz) are read by moving the marker over the spectrogram. The intensity level (on the left of the figures, the color scale) of the marked and nonmarked frequency limits of the sounds is between 45 and 50 dB (yellow color in the figures).

Hence, this system is able to display and reproduce cardiovascular sound events. For example, a musical murmur caused by harmonic movements of the heart and vasculature, such as an innocent vibratory murmur (Still's murmur) is usually visualized as a well-defined area or line on the spectrogram. Innocent systolic murmurs appear to have a lower peak frequency (below 200 Hz) and a shorter duration (below 80% of the duration of "audible systole" from the end of S1 to the beginning of S2) compared with recordings of pathologic murmurs. Also, innocent systolic murmurs always fade before the second heart sound. In contrast, the higher velocity blood flow of a pathologic cardiac lesion produces a more intense murmur and a wider frequency scale (in hertz). This phenomenon is clearly demonstrated on the spectrographic recordings.

Limitations of Phonocardiography

Background noise can be eliminated with filtering systems, but phonocardiography still has limitations. The intensity of heart sounds varies between different auscultation areas and different persons. The loudness of heart sounds (in dB) is difficult to illustrate with the phonocardiographic method. Spectrograms may give more reliable information about the quality of a heart murmur, as seen in the examples in Figs. 2–4, 2–5, 2–6, 2–7, 2–8, 2–9, 2–10, 2–11.

Heart Sound Analysis in Telemedicine

Telemedicine is a rapidly developing application area in clinical medicine. It allows the specialist to work in a wider geographical area and to offer services in rural areas. Today, data can be conveniently stored on personal computers or hand-held devices and transferred through communication networks, like the Internet or mobile phone networks, assuming such networks are available. Heart sound signals can be remotely monitored in real time in teleconsultations, the examination can be done interactively, or the signals can be stored and forwarded for more detailed analysis and consultation purposes.19-21

Clinical Auscultation

The acoustic information of phonocardiography visualizes the auditory events of the heart. Spectral phonocardiography emulates the ear and can be used during a teaching session. In clinical practice, observation of spectrograms or digital analysis of recorded heart sounds compensates for the inexperience of the listener. However, it must be kept in mind that this method is only a part of the clinical examination. It should never remain as the only criterion of a child's cardiac defect. The method may eliminate unnecessary consultations or echo examinations, but it should never replace echo examinations.

Fetal Heart Sound Doppler Recording

Fetal heart rate monitoring using the Doppler shift resulting from the movements of the fetal heart is a standard examination in most obstetrical wards. Ultrasound-based instruments are not suitable for long-term surveillance of the unborn. Heart sound monitoring with an electronic stethoscope is noninvasive, and digital phonocardiography has proved to be a useful method for prenatal heart sound follow-up and screening of fetal heart murmurs.22-24

Laser Doppler Vibrometry

Laser Doppler vibrometry has been recently applied to noncontact monitoring of cardiac activity, both in terms of cardiac rate and heart rate variability, by measuring the velocity of the skin surface of the chest wall and the neck (optical vibrocardiography). The data were compared with phonocardiography, and the findings were correlated with the heart sounds related to the closure of the mitral valve and the following closure of the aortic and pulmonary valves with characteristic deflections identifiable on the vibrocardiography traces.25,26

Electrocardiography and Phonocardiography

In clinical practice or during a teaching session, one sound channel without an electrocardiogram (ECG) is a fast and easy way to record and store sound signals. However, a simultaneous ECG recording is advisable for accurate timing and documentation of the cardiac cycle.

Assessment of the left ventricular systolic function has been applied to derive the systolic time intervals. The method combines analogic phonocardiography with ECG and external recordings of the carotid pulse. More precise noninvasive assessments of the left ventricular systolic function are performed with acoustic cardiography—a simultaneous ECG and heart sound recording with digital analysis.15-18

Pulse Doppler echocardiography is the most common method for predicting optimal atrioventricular synchrony in patients with a complete atrioventricular block and an implanted DDD pacemaker. The possibility of using phonocardiography instead of echocardiography has been demonstrated to be a valid and simple method for this purpose.27

Cardiac Catheterization and Phonocardiography

During radiofrequency ablation, the acoustic energy created by microbubble formation can be detected without echocardiography using digital phonocardiography.28

Echocardiography and Phonocardiography

Simultaneous timing of heart sounds is used with echocardiography (eg, by assessing the ventricular function). Noise and filtering may limit the precision of phonocardiography.29

Magnetic Resonance Imaging and Phonocardiography

A novel device, the magnetic resonance stethoscope, consists of an acoustic sensor, a signal processing unit, and a coupler to the magnetic resonance imaging system. The system is safe, there is no risk of high-voltage induction or patient burns, it is immune to electromagnetic interference, and it is suitable for all magnetic fields.30

During the past years, improvements in personal computers have made it possible to design new low-cost, high-quality phonocardiographic devices. Spectral phonocardiography enhances auscultation and is a valid and valuable tool in clinical practice. It can be used to teach an inexperienced listener the art of auscultation. The computerized signal processing method enables automatic analysis of audible heart events and thereby offers a new diagnostic aid in clinical decision making. Digital phonocardiography provides the technology to accurately record and store cardiac auscultatory findings, thus enabling store-and-forward teleconsultation and improved documentation in electronic medical records. The digital processing of heart sounds allows the combination of phonocardiography with other imaging modalities.