International Journal of Scientific & Engineering Research, Volume 2, Issue 2, February-2011 1

ISSN 2229-5518

An Effiecient Modeling Technique for Heart

Sounds and Murmurs

Kiran Kumari Patil*, Dr.B.S Nagabhushan, Dr. Vijaya Kumar B.P

(*Research Scholar Dr. MGR University, Chennai)

Abstract— Cardiac auscultation is highly subjective and a cognitive process and the amount information that can be obtained by listening heart sounds largely depends on the expertise, experience and acuity of the ear of the physician. In general, the classification and interpretation of heart sounds and murmurs is based on a adjective 0-6/6 grade scale and described by using “faint”, “soft”, “loud”, “ high pitch”, “clear”, “thrill”, “tremor”, “musical” and others terms. These terms are not well-defined and suitable mathematical models are not available. Apart from that, the adjective scales vary among the doctors and difficult to derive a standard model for heart sound quality and correct clinical interpretation.

In this research paper, we propose a novel framework based on psychoacoustic principles and derive psychoacoustic models for the heart sounds and murmurs. We discuss the theoretical foundations, psychoacoustic principles and derive the mathe- matical models for the psychoacoustic features such as loudness, sharpness, intensity, strength, roughness, tonality etc. for a set of heart sounds and murmurs. The proposed framework helps in deriving heart sound quality and also used for the com- parison and correlation with normal and pathologic murmurs and enhances clinical decisions.

.

Index Terms— psychoacoutic principles, psychoacoutic models, heart sounds, auscultation, phonocardiography

.

1 INTRODUCTION

—————————— • ——————————

Heart sounds and murmurs are acoustic phenomenon caused by the mechnical events of cardiac system. Aus- cultation or hearing of heart sounds using conventional stethscope or electronic stehoscope is not purely a mech- nical phenomenon of sound wave propagation, but also auditory, sensory, cognative and perceptual event. Digital cardiac auscultation is an art and science of interpreating hearts sounds and murmurs used for clinical diagnostics. Congnitive process that paly significant role heart sound perception and clinical interpreatation. The phonocardio- graphy (PCG) — the art and science of recording and in- terpreting of heart sounds using latest digital technology has significantly helped to understand and interpret the complex heart sounds (normal, abnormal sounds includ- ing murmurs). The PCG is a display of the heart sound signal with respect time (time domain) and frequency components or spectral properties (frequency domain) and plotting the heart sounds and murmurs can provide useful information to the physician by complementing cardiac auscultation Phonocardiography techniques are used for the effective clinical investigations and corrective diagnostic heart related diseases and in particular valvu- lar heart diseases. When a heart valve is stenotic or dam- aged, the abnormal blood flow patterns produce a series of audible vibratory sounds known as murmurs [3]. Doc-
tors and physicians detect these disorders by listening to
heart sounds at different locations across the torso. Mur- murs heared during routine physical exminations offer important clues to the presence of undetected and asymp- tomatic cardiac disease. The process of interpreating heart sounds is called cardiac auscultation. It is a simple, non- invasive tec niques helps in early detection of cardiac dis- orders. Different cardiac ailments produce a potentially overwhelming set of acoustic pathological events and correctly identfying a disorder requires discrimination of subtle variations in the timing characteristics and spectral properties of heart sounds. The analysis and interpreta- tion is complicated by the natural variations in heart sounds introduced by factors such as the gender, age, habitus and dynamic state of the patients. Even in adults, anixiety, stress, fever, anemia etc. may also cause benign murmurs. Typically, these cases are distinguished by ex- amining the intensity of sounds, in addition to their tim- ing and frequency content. Sounds that are interesting from the perspective of auscultation are often short lived (less than 20 millseconds) and seperated from one another by less than 30 milliseconds. Pathological signals indica- tive of cardiac diseases are also often much quieter than other heart sounds and their audibility varies across su- cessive heart beats. Even with extensive experience, phy- sicians may often disagree about sounds, in particular
with brief heart sounds. These inaccuracies are attributed
IJSER © 2011 http://www.ijser.org

2 International Journal of Scientific & Engineering Research, Volume 2, Issue 2, February -2011

ISSN 2229-5518

to human auditory limitations which include insensitivity to frequencies, slow response to rapidly occurring changes in acoustic signals and an inability to unmask soft sounds in the promixity of loud ones. Murmurs are extra heart sounds that are produced as a result of turbu- lent blood flow which is sufficient to produce audible noise. Murmurs may be present in normal hearts without any heart diseases are called innocent murmurs. The murmurs due to valvular heart diseases are called patho- logic murmurs and needs to evaluate by the cardiologists. In general, the murmurs [3, 4] can be classified by seven different physical characteristics: timing, shape, location, radiation, intensity, pitch and quality. It may also include psychoacuostic characteristics such as loudness, heart sound intensity, heart sound pressure, sharpness and fluctuations. Timing refers to whether the murmur is a systolic or diastolic murmur depending on the S1 and S2 sounds of a standard heart sound cycle and plays vital role in clinical decisions.

Shape refers to the intensity of the heart sounds over time. We can derive the intensity contours and can be de- scribed in musical note such as “cresccendo”, “decrescen- do”, or “crescendodecrescendo”.

Radiation refers to where the sound of the heart sounds and murmurs radiates and normally radiates in the direc- tion of the blood flow. The radiated sound can be cap- tured and displayed as radiation patterns and helps in specific characterizations of murmurs.

Intensity refers to the loudness of the murmur, and is graded on a scale from 0-6/6 adjective scale as shown in the following Table I. The terms such as “soft”, “faint”, “loud” ,”blowing”, “harsh”, “rumbling”and others have a subjective meaning and needs to be modeled precisely

Pitch of a murmur is low, medium or high and is de- termined by whether it can be auscultated best with the bell or diaphragm of a stetchscope. Pitch mainly refers to the fundmental frequency of the heart sounds and murmurs.

From the above discussios, it clear that the physical cha-
racteristics such as timing, shape, pitch, loudness, radia- tion, sound intensity, sound pressure and other parame- ters are clinically important and and described using a adjective scales as shown in table 1.

Table 1: Gradings on a scale from 0-6/6

In order to address the above challenges, we derive psy- choacoustics principlaes for the hearts sounds and mur- murs in a novel way and specfic contributions of the re- search work. We propose and derive psychoacoutsic models based on the psychoacoustics principles for heart sounds and murmurs.We derive psychoacoutsic models and map them to the psychoacustic features such as loundness, pitch, sound intensity, sharpness, etc. with mathematical equations.We also discuss the classifciation of murmurs and adjective scales and also try to relate them into the psychoacoustic features. For example, the grade 4 murmur which is loud and thrill and derive the loundness contour and characterize the thrill by observ- ing the contour of loudness pattern.We highlight the usage of psychoacuostic model for the heart sound quali- ties.

2. PSYCHOACOUSTIC PRINCIPLES AND MODLES

Psychoacoustics is the study of the subjective human per- ception of sounds [1]. Alternatively it can be described as the study of psychological correlates of the physical pa- rameters of acoustics [1]. The field of psychoacoustic aims to model parameters of auditory sensation in terms of physical signal parameters and provide a framework and modeling capabilities for the acoustic sounds. The psy- choacoustic models of sound perception exploiting the imperceptable sounds are used in the audio compression such as MP3 standards; non-linear response of the ear is exploited in the noise reduction systems and communica- tion networks. The human ear can nominally hear sounds
in the range 20 HZ to 20,000 Hz (20 kHz). Frequency reso

IJSER © 2011 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 2, Issue 2, February-2011 3

ISSN 2229-5518

lution of the ear is 0.36 Hz within the octave of 1,000–
2,000 Hz. That is, changes in pitch larger than 0.36 Hz can be perceived in a clinical setting. Other scales have been derived directly from experiments on human hearing perception, such as the Mel scale and Bark scale and these are approximately logarithmic in frequency at the high- frequency end, but nearly linear at the low-frequency end. Ear drums are sensitive only to variations in the sound pressure, but can detect pressure changes as small as 2×10–10 ATM and as great or greater than 1 ATM. The sound pressue level (SPL) is also measured logarithimi- cally, with all pressures referenced to 1.97385×10–10
ATM. The lower limit of audibility is therefore defined as
0 dB, but the upper limit is not as clearly defined. By measuring this minimum intensity for testing tones of various frequencies, a frequency dependent absolute threshold of hearing (ATH) curve may be derived. Typi- cally, ear shows a peak of sensitivity (i.e., its lowest ATH) between 1 kHz and 5 kHz, though the threshold changes with age, with older ears showing decreased sensitivity above 2 kHz. Equal-loudness contours indicate the sound pressure level (dB), over the range of audible frequencies, which are perceived as being of equal loudness and may be ploted. We use classical reference [5] which represents a set of algorithms for calculating auditory sensations including loudness, sharpness, roughness, softness, sound strength and intensity, and fluctuation strength and extend it for the heart sounds and murmurs. The classification of murmurs is characterized by using the psychoacoustic features and derives mathematical equa- tions.

3. PROPOSED MODELS

3.1 Loudness of heart sound and murmurs

The loudness is modeled by the following equation, where N is loudness, N’ is the loudness of a given critical band or also know as specific loudness, and dz is the in- crement in the critical band scale.

24

murmurs in clinical investigations. We can plot the loud- ness contour using above equations and derive deep in- sight in the heart sound analysis and interpretations. The complexity of the loudness of complex heart sound needs to be investigated in multistages and becomes compli- cated for time varying sounds, dynamical effects like for- ward masking and temporal loudness integration [2, 6, 5] have to be considered When the physician hears the heart sounds and murmurs, the loudness is a clinically signifi- cant feature and can be measured in sone and be mapped to the vertical axis of the Figure 1 in terms of the adjective scales of murmur classifications.

Figure 1. Loudness function of a 1- kHz tone (solid) and of Uniform Exciting Noise (dotted), broken and dashed-dotted lines with cor- responding equations (adopted from [5])

3.2 Sharpness of heat sound and murmurs

Sharpness or brightness is one of the most prominent fea- tures of timbre. Timbre is more complicated, being de- termined by the harmonic content of the signal. The hear- ing is based on the amplitude of the frequencies and is very insensitive to their phases. The shape of hearts sounds and murmurs in time domain waveform is only indirectly related to hearing and poses serious challenges in correct interpretation of heart sounds. The models are based on the centroid (signal spectrum or loudness pat- tern) of the heart sounds and murmurs. The sharpness is modeled as a weighted centroid of the specific loudness pattern. The unit is acum, referenced to a band of noise 1 critical band wide, centered on 1 kHz at 60 dB. It is also
referred to the perception that the sound is “sharp”,

N = f N 'dz

0

(1)
“harsh” or “soft” when used in the context of heart sound perception and clinical interpretations. It is related to the
The unit of loudness, the sone, is a ratio scale referenced against the sensation produced by a 1 kHz sine tone with a sound pressure level of 40 dB. The models draw on data gained from subjective testing and from a physiological understanding of the auditory periphery. The “loudness” property of heart sounds and murmurs helps in characte-
rization, classification and discrimination of various
proportion of high frequency energy present in the
sound, weighted towards energy in the region above 3 kHz. For harmonic tones, sharpness can be controlled through distribution of the harmonic spectral envelope. The model used [5, 9, 7] for the calculating the sharpness of tones is summarized by the Equation 2.

IJSER © 2011 http://www.ijser.org

4 International Journal of Scientific & Engineering Research, Volume 2, Issue 2, February -2011

ISSN 2229-5518

24 Bark

f

N ' g ( z ) zdz

sound is heard best with the diaphragm of the stethos- cope, i.e., high pitched, or with the bell, i.e., low pitched.

s = 0.11 0

f N 'dz

0

(2)
Murmurs of mitral or tricuspid stenosis are best heard
with the bell. Some of other hearts sounds that can cha- racterize the pitch are: S3 sound is a low frequency, mid
where the S is sharpness, N’ is specific loudness, z is the bark scale of auditory filters and g(z) is a weighting func- tion that emphasis z for the critical band rates. It was found that the sharpness of narrow band noises increases proportionally with the critical band rate for center fre- quencies below about 3 kHz. At higher frequencies, how- ever, sharpness increases more strongly, an effect that has to be taken into account when the sharpness S is calcu- lated using a formula that gives the weighted first mo- mentum of the specific loudness pattern: In equation (2), the denominator gives the total loudness, while the upper integral is the weighted momentum mentioned. The psy- choacoustic feature - sharpness can be used for the fol- lowing murmurs.
Early systolic ejection click murmur is a high frequency, early systolic sound occurring 0.03- 0.07 second after S1 [4]. The sound is generated either by the sudden upward doming of an abnormal semilunar valve (aortic or pluton- ic) and sharp click sounds.
Opening Snap murmur is a high frequency, early diastolic sound that is associated with MS [3]. It occurs 0.04-0.12 second after S2 and may or may not be associated with a late peaking or rumbling diastolic murmur of varying sharpness with peaking at specific temporal patterns.

3.3 Pitch of heart and murmurs

A psycho acoustical pitch ratio scale is a difficult concept due to the complexity of pitch perception and cognition. Reference [6, 10] describes some of the complexity of pitch structures for harmonic tones (such as pitch height,
octave equivalence and cycle of fifths) through multidi-
diastolic sound occurring 0.14-0.22 second after S2. The
frequency components of low frequency heart sounds are difficult to hear and can be modeled and uniquely find pitch features of the heart sounds and murmurs [3]. S4 sound is also a low frequency, late diastolic sound occur-
ring 0.08-0.20 second prior to S1. It is generated during
presystolic ventricular filling due to atrial contraction, hypertension and diastolic dysfunction. The sequencing and ordering with respect to the S1 and S2 in a standard cardiac cycle and obtain pitch pattern and frequency components using spectral techniques will assist the doc- tor for the better clinical decisions [3].

3.4 Roughness of heart sounds and murmurs

Roughness is a sensation caused by quite rapid amplitude modulation within auditory filters. This modulation can be caused by beats between two pure tone components, or by a signal with amplitude or frequency modulation. Beating within an auditory filter channel has been used to explain the acoustic component of tonal dissonance and represent roughness of the heart sounds [5]. The unit of roughness is the asper, which is referenced to a 1 kHz tone at 60 dB with 100% amplitude modulation at 70 Hz. The model presented in [5,8] by for calculating the roughness of modulated tones having a single modula- tion frequency is given in Equation 5, where R is rough- ness, fmod is the modulation frequency, and LE is the ex- citation level within an auditory filter. This uses the time- varying excitation pattern of the ear (similar to the specif- ic loudness pattern, except that the magnitude is in deci- bels rather than sones/bark), with the difference between maximum and minimum excitation levels integrated across auditory filters used to determine roughness.

24 Bark

mensional geometric figures. In addition to complex

R = 0.3 fmod

f ').LE ( z)dz asper

(3)

structures of pitch height, pitch has the dimension of
pitch strength, also known as “tonalness”. The harmonic series is of great importance in pitch perception, and mainly pitched sounds in everyday experience exhibit harmonic spectra. In general, it is usually determined by the fundamental frequency as a pitch percept. A model of pitch perception is analyzed using template matching or autocorrelation techniques. Pitch is used to describe the tonal quality of the murmur be it high pitched or low pitched. For those of us not musically inclined, a simple
way to distinguish pitch is to determine whether the

kHz 0 dB / Bark

The roughness of the heart sound can be used for the modeling aortic insufficiency (AI) may be congenital rheumatic, and collagen vascular disease. The murmur is a high frequency (blowing) decrescendo murmur begin- ning in early cardiac cycle and uniquely radiates to the top of the head. The roughness of the murmurs can be uniquely characterized using the above equation and needs further investigations. Mitral regurgitation (MR) is

IJSER © 2011 http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 2, Issue 2, February-2011 5

ISSN 2229-5518

associated with endocarditis, and ischemic heart diseases [4]. The murmur is typically a high frequency, holosystol- ic, plateau murmur that is best heard at the apex. The murmur often radiates to the left axilla and back. There is no appreciable change in murmur intensity with cycle length (as with AS). MR may be associated with S3 in more severe cases. Here we have to use intensity, high frequency and radiation patterns in a consistent way and need further investigation and clinical validations.

4. CONCLUSIONS

The cardiac auscultation is an effective diagnostic tech- nique used in the early detection of cardiac diseases in particular the valvular diseases, including the murmurs. It is argued that the most of the doctors and physicians depend on their experience and make subjective interpre- tations of heart diseases. In this paper, we proposed psy- choacoustic models based on a psychoacoustic principles and mathematical foundations and discussed the psy- choacoustic features (pitch, intensity, timbre, loudness, power, intensity and other clinically important psychoac- oustic features) that can be modeled, analyzed and pro- vide effective aid of clinical decisions related to heart dis- eases, and in particular murmurs. These models offer a reasoning framework for the subjective reasoning of heart sounds and derived psychoacoustical models. It is also used to model the quality of heart sounds for many stan- dardization efforts and can be used as an effective teach- ing aid for the cardiac auscultations. Our preliminary investigations and experimental results on our psychoac- oustic models are quite encouraging and provide a dee- per insight into the perception and interpretation of car- diac auscultations. The visualization tools for the psy- choacoustic models are in progress and will help in clini- cal decisions. Further investigations and validation of the proposed psychoacoustic models are planned for the fu- ture work.

ACKNOWLEDGMENT

We thank Dr. R.P Reddy, Principal, REVA College ofEn- gineering, Bangalore for his constant support and motiva- tion for the research work. We thank Prof. Dr. Cyril Raj, MGR University, Chennai, for research discussions, for providing critical inputs and guidance for the research work.

REFERENCES

[1] J. P. Christopher, The Sense of Hearing, 2nd ed., Routledge, New

York,2005.

[2] J. Katz, R.F. Burkard and L. Medwetsky, Handbook of Clinical

Audiology, Lippincott Williams and Wilkins, 2002, New York.

[3] P.Stein, A physical and physilogical basis for the interpreation of cardiac auscultation: evaluations based primarily on the second sound and ejection murmurs, Futura Publishing Company, 1981.

[4] B. Shipton, W. Wahba, “Valvular heart disease: review and up- date”, American Family Physician, 63(11), pp. 2201-2208, 2001

[5] E. Zwicker and H. Fastl, Psychoacoustics: Facts and Models, Springer, Berlin, 1999.

[6] B.C.J More, B.R. Glasberg and T. Baer, “A model for the predic- tion of thresholds, loudness and partial loudness”, Journal of the Audio Engineering Society, vol. 45, no. 4, pp. 224-240, 1997.

[7] B.R Glasberg and B.C Moore, “Amodel of loudness applicable to timevarying sounds”, Journal of the Audio Engineering Society, Vol.

50, No. 5, pp. 331-342, 2002

[8] P. Daniel and R. Weber, “Psychoacoustical roughness: impleman- tion of an optimized model”, Acustica, no. 83, pp. 113-123, 1997.

[9] G.V Bismark, “Sharpness as an attribute of the timbre of steady sounds”, Acustica, vol. 30, pp. 159-172, 1974

[10] S.S Stevens, J. Volkman and E. Newman, “A scale for the mea- surement of psychological magnitude of pitch”, Journal of Acoustic- al Society of merica, vol. 8, pp. 185-190, 1937.

IJSER © 2011 http://www.ijser.org