International Journal of Cardiology 172 (2014) 548–560

Contents lists available at ScienceDirect

International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard

Review

Beyond auscultation: Acoustic cardiography in clinical practice Yong-Na Wen 1, Alex Pui-Wai Lee 1, Fang Fang 1, Chun-Na Jin 1, Cheuk-Man Yu ⁎,1 Division of Cardiology, Department of Medicine and Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong

a r t i c l e

i n f o

Article history: Received 23 December 2013 Accepted 30 December 2013 Available online 10 January 2014 Keywords: Acoustic cardiography AUDICOR Phonocardiography

a b s t r a c t Cardiac auscultation by stethoscope is widely used but limited by low sensitivity and accuracy. Phonocardiogram was developed in an attempt to provide quantitative and qualitative information of heart sounds and murmurs by transforming acoustic signal into visual wavelet. Although phonocardiogram provides objective heart sound information and holds diagnostic potentials of different heart problems, its examination procedure is timeconsuming and it requires specially trained technicians to operate the device. Acoustic cardiography (AUDICOR, Inovise Medical, Inc., Portland, OR, USA) is a major recent advance in the evolution of cardiac auscultation technology. The technique is more efficient and less operator-dependent. It synchronizes cardiac auscultation with ECG recording and provides a comprehensive assessment of both mechanical and electronic function of the heart. The application of acoustic cardiography is far beyond auscultation only. It generates various parameters which have been proven to correlate with gold standards in heart failure diagnosis and ischemic heart disease detection. Its application can be extended to other diseases, including LV hypertrophy, constrictive pericarditis, sleep apnea and ventricular fibrillation. The newly developed ambulatory acoustic cardiography is potentially used in heart failure follow-up in both home and hospital setting. This review comprehensively summarizes acoustic cardiographic research, including the most recent development. © 2014 Elsevier Ireland Ltd. All rights reserved.

1. History of cardiac auscultation Before the 19th century, cardiac physical examination was limited to inspection and palpation. With the invention of stethoscope by Rene Laennec in 1816, cardiac auscultation was introduced into clinical practice for cardiovascular examination [1]. For two centuries, physicians have been using this method to interpret physiological and pathological changes of the heart. However, it has been reported that auscultation by stethoscope has low diagnostic sensitivity and accuracy [2,3]. Clinical disagreement and skill discrepancy among examiners result in non-objective and unreliable auscultation data [4,5]. As reported by Avendano-Valencia et al., human ear limitation is another explanation to the deficiency of traditional physical auscultation [6]. Phonocardiogram was developed in an attempt to provide quantitative and qualitative information of heart sounds and murmurs. The phonocardiogram records heart sound through a microphone, which is placed on the chest wall of subjects. The acoustic signal is transformed into visual wavelet and recorded graphically [7]. Mathematical methods to improve the diagnostic quality of phonocardiogram data have evolved in recent years [3,8–10]. With the advance of digital techniques, ⁎ Corresponding author at: Institute of Vascular Medicine (IVM), Li Ka Shing Institute of Health Sciences (LiHS), S.H. Ho Cardiovascular Disease and Stroke Centre, Heart Education And Research Training (HEART) Centre and Division of Cardiology, Department of Medicine and Therapeutics, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong. Tel.: +852 2632 3594; fax: +852 2637 5643. E-mail address: [email protected] (C.-M. Yu). 1 All authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation. 0167-5273/$ – see front matter © 2014 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijcard.2013.12.298

phonocardiogram is now capable of characterizing heart sounds and murmurs by their acoustic components, pattern, timing, frequency and intensity [2]. Data obtained from phonocardiogram can refine cardiologists' understanding of the mechanism and origin of heart sounds and may aid in clinical diagnosis. The results of phonocardiogram have been used as reference standard to judge physicians' auscultation skill [11]. The application of phonocardiogram was further extended to the diagnostic evaluation of left ventricular (LV) dysfunction and myocardial ischemia. Systolic time interval, i.e., the interval between peak intensity of the first (S1) and second heart sounds (S2), has been shown to correlate with impaired LV systolic function and coronary heart disease [12–21]. Phonocardiogram, when combined with echocardiography, electrocardiography (ECG) and carotid pulse tracing offers a non-invasive measure of systolic time intervals to evaluate LV function [12,16,18,19] and to detect coronary artery insufficiency [14,17,20,21]. Although phonocardiogram provides objective heart sound information and holds diagnostic potentials of different heart problems, its examination procedure is time-consuming and requires specially trained technicians to operate the device. Consequently, a more efficient and less operator-dependent technique is needed for routine clinical practice, and as a result, acoustic cardiography (Audicor, Inovise Medical, Inc., Portland, OR) was developed.

2. The technology of acoustic cardiography Acoustic cardiography is a major recent advance in the evolution of cardiac auscultation technology. The technique synchronizes cardiac

Y.-N. Wen et al. / International Journal of Cardiology 172 (2014) 548–560

auscultation with ECG recording and provides a comprehensive assessment of both mechanical and electronic function of the heart. The Audicor system consists of three components: an acoustic/ECG console with two chest sensors, a computer for signal processing and analysis, and a printer to generate a full report of acoustic and ECG data. The acoustic information is processed with wavelet signal reprocessing techniques and uses time-frequency analysis to detect heart sound, as well as the intervals between sounds (Figs. 1–7). Unlike traditional phonocardiogram, acoustic cardiography has the advantage of requiring minimal training in its operation. The examination procedure is even simpler and faster than a traditional 12-lead ECG recording. By placing two sensors on the V3 and V4 positions of the chest and attaching 2 standard electrodes on limb leads positions of standard ECG, both the heart sounds and ECG data are captured, algorithmically analyzed, stored, and displayed. Traditional phonocardiogram required trained clinicians or experienced technicians to interpret the acoustic signals, increasing its inter- and intraobserver variability. In contrast, in acoustic cardiography a detailed report of computerized analysis of the acoustic and ECG information is automatically generated within seconds after data acquisition (Fig. 8), simplifying clinical workflow. The latest advent of acoustic cardiography is the development of ambulatory monitoring technology (Audicor AM) to allow continuous surveillance of acoustic and ECG data for detection of infrequent events such as transient ischemia, ventricular dysfunction, arrhythmia as well as sleep apnea. The Audicor AM device and data collection process are similar to that of conventional Holter monitoring. Signal acquisition is achieved by attaching 3 electrodes to the chest wall of patients and having the patients to carry a small recorder unit for a period of 24–48 h. Through a wireless connection to a computer, data can be previewed while on recording. A removable smart card within the recorder unit is used to save up to 48 h of data. Data is then transferred to computer for algorithmic analysis. Compared with snapshot acoustic cardiogram, which records only isolated events, ambulatory monitoring device shows the trending of heart failure and therefore evaluates the therapeutic response. Ambulatory acoustic cardiography not only provides acoustic and electrical information, but also respiratory events. With another automated algorithm, the ambulatory recorder unit simultaneously acquires respiratory data, which includes respiration rate, episodes of snoring and sleep apnea [22] (Fig. 9). While a 3-second snapshot acoustic cardiography can be too brief to capture infrequent events, the Audicor AM ambulatory recording device may allow detection of silent ischemia, LV dysfunction, filling status, responses to therapeutic measures and sleep apnea by allowing continuous out-of-hospital monitoring.

549

3. Parameters of acoustic cardiography Acoustic cardiography generates a number of parameters in connection of p wave and QRS complex in every cardiac cycle from ECG to represent LV systolic and diastolic functions (Table 1 and Fig. 10). Among them, the diagnostic value of EMAT, SDI, S3 and S4 in heart failure have been validated in clinical studies and their cut-off values are also established (Table 2) (Fig. 11). 3.1. EMAT and %EMAT EMAT represents the time required to generate force by LV to close mitral valve and is therefore related to the acceleration of the pressure in LV. Previous studies have indicated that: - Longer EMAT or increased %EMAT indicates impaired LV function [23]; - Shorten EMAT correlated with improved LV contractility and short electromechanical delays [23]; - Increased %EMAT predicted re-hospitalization in heart failure [24]. 3.2. SDI (QRS duration ∗ QR interval ∗ %EMAT ∗ S3 strength) SDI is a multiplicative combination of ECG and acoustic parameters to predict LV systolic dysfunction with high specificity. It has been shown that: - SDIN = 5 suggested LV systolic dysfunction (EF b 50%); - SDI N 7.5 suggested severe LV systolic dysfunction (EF b35%) [25]. 3.3. S3 Acoustic cardiography objectively and sensitively detects S3 and expresses it in a value of 0 to 10. S3 strength ≧5 declares the presence of S3. In patients over 40 years, S3 is considered pathological and associated with elevated LV filling pressure and impaired LV contractility. Utilizing acoustic cardiography, studies show that: - S3 had a positive likelihood ratio of 4.8 to predict LV dysfunction [27]; - S3 correlated with increased LV end-diastolic pressure [27,29]; - S3 assisted BNP to increase diagnostic accuracy of acute heart failure [26]. 3.4. S4 Similar to S3, acoustic cardiography derived S4 strength is also provided in a value of 0 to 10. Presence of S4 (≧5) is generally pathological and suggests increased LV stiffness. It was shown that - S4 was associated with LV stiffness and elevated LVEDP [27]. - The nocturnal increase in S4 strength in asymptomatic older patients indicates diastolic impairment consistent with echocardiographic diastolic filling patterns with increasing age [49]. 4. Clinical application of acoustic cardiography 4.1. Heart failure diagnosis and monitoring

Fig. 1. 3D sound fingerprint. This figure shows the heart sound analysis with wavelet techniques, resynchronized with ECG. Each heart sound has its own time, frequency and energy distribution.

4.1.1. Evaluation of left ventricular function Acoustic cardiography has been shown to correlate with invasive and non-invasive cardiac hemodynamic assessment of LV function. Previous research suggested that acoustic cardiography parameters detects LV dysfunction as precisely as echocardiographic parameters, including LVEF (28,29)and LVDP [30]. In a study of 161 heart failure patients conducted by Zuber et. al., EMAT, LVST, EMAT/LVST measured by acoustic cardiography strongly agreed with echocardiographic ejection fraction (EF) measurement to identify LV systolic dysfunction [28]. In a subsequent larger scale study in 433 heart failure patients, Kosmicki

550

Y.-N. Wen et al. / International Journal of Cardiology 172 (2014) 548–560

Fig. 2. Acoustic cardiography report of diastolic heart sound and systolic murmurs. The upper graph is the example of acoustic cardiography report, which shows detection of detection of diastolic heart sound in consistent with LV dysfunction, and suspicious of LBBB. The below figure is a time-frequency analysis of heart sound and murmurs.

et al. demonstrated consistently that Audicor-derived EMAT and LVST are strong predictors of heart failure with reduced ejection fraction (HFREF), using echocardiography as the reference. In addition, the study also showed that acoustic cardiography predicted depressed LVEF more accurately than BNP did (sensitivity 88% vs 80%, specificity 71% vs 55%, positive likelihood ratio 3 vs 1.8) [29]. Acoustic cardiography was also shown to correlate strongly with invasive techniques in the measurement of LV dp/dt [30] and pulmonary capillary wedge pressure [31]. Using catheter-based measurement of LV dp/dt b1600 as definition of LV systolic dysfunction, Markus R. et al. observed in 108 patients that EMAT generated by acoustic cardiography was significantly prolonged in patients with reduced LV dp/dt. Our group was the first to validate the clinical utility of acoustic cardiography in differentiation of heart failure phenotypes. In 3 groups of patients, including 94 with hypertension, 109 with HF and normal ejection fraction (HFNEF, EF ≥ 50%) and

89 with HF and reduced ejection fraction (HFREF, EF b 50%), we investigated the discriminating power of acoustic cardiography to diagnose heart failure and to differentiate HFNEF from HFREF. We demonstrated that EMAT/RR significantly differentiated HFNEF from hypertension (AUC = 0.83) with an EMAT/RR N 11.54% yielded 55% sensitivity and 90% specificity, similar to the diagnostic performance of echo-measured E/e′ N 15 (AUC = 0.84, 55% sensitivity, 90%). Moreover, our results showed that SDI out-performed other acoustic cardiographic parameters in differentiating HFREF from HFNEF (AUC = 0.81), and an SDI N 5.43 yielded 53% sensitivity and 91% specificity. The E/e′ ratio had a similar diagnostic performance. The Audicor may therefore be a promising bedside tool to provide rapid and accurate differentiation between HFand non-HF-related causes of dyspnea, and may help identifying HF phenotypes, especially when echocardiography is not immediately available [32].

Y.-N. Wen et al. / International Journal of Cardiology 172 (2014) 548–560

551

Fig. 3. Pericardial knock. (a) An acoustic cardiography suggests ischemia heart diseases. (b) A 2D and 3D time-frequency analysis shows pericardial knock noted, which occurs at the same time as S3 but has higher frequency components.

Most recently, acoustic cardiography was validated in monitoring and early detection of LV dysfunction in cancer patients exposed to. It was suggested that lower baseline LVEF was predictive to later developed LV systolic dysfunction after chemotherapy. However, under current guideline, there is no sufficient evidence to support usefulness of baseline echocardiographic screening of LV function for all patients receiving chemotherapy. Toggweiler et al. prospectively observed 187 cancer patients treated with anthracyclines and performed acoustic cardiography right after the therapy. They found patients %EMAT N12.4% and EMAT N95 performed post chemotheraphy had very high sensitivity (88% and 75%) and specificity (both 84%) to identify LV systolic

dysfunction as confirmed by echocardiography [33]. As of such, while performing echocardiography on all of the cancer patients may be unnecessary, acoustic cardiography may assist in selecting high risk patients for further echocardiographic examination. 4.1.2. Detection of the third heart sound and clinical diagnosis of heart failure It has long been recognized that the presence of S3 is a sign of significant LV dysfunction and acute decompensated heart failure. Previous studies demonstrated that S3 had a positive likelihood ratio of 4.8 to predict LV failure [34], correlated with depressed LVEF, increased E

552

Y.-N. Wen et al. / International Journal of Cardiology 172 (2014) 548–560

Fig. 4. Artificial aortic valve. (a) An acoustic cardiography report of a subject with prosthetic aortic valve shows no abnormality. (b) Time-frequency shows high frequency of aortic valve sounds, with both opening and closure audible.

deceleration, E/e′ and filling pressures [32,35–38]. In addition to its diagnostic value, S3 also carries important prognostic information. It was reported that a clinically detectable S3 was very predictive of future adverse cardiac events, including death and HF-related re-hospitalization, compared with those without [39–41]. However, the sensitivity and accuracy in detection of S3 by traditional stethoscopes are limited due to drawbacks of physical auscultation as discussed above. Co-morbidities of some patients such as COPD and obesity make S3 auscultation even more problematic [42]. Acoustic cardiography provides a possible

solution to these problems by digitally recording and quantifying S3 in an objective range of 0–10. Presence of S3 is defined as a value equal or over 5. By viewing acoustic cardiography, even experienced clinicians could significantly improve their S3 detection accuracy by 8–18% [43]. Moreover, acoustic cardiography helps to correct 34% of the patients initially missed for diagnosis of acute heart failure [44]. The clinical value of acoustic cardiography in providing rapid bedside diagnosis for dyspneic patients presenting to the emergency

Y.-N. Wen et al. / International Journal of Cardiology 172 (2014) 548–560

553

Fig. 5. Phrenic nerve stimulation. (a) An acoustic cardiography report shows a pace-maker rhythm. (b) Time frequency detects diaphragmatic sound, which is high in frequency and occurs 20–30 ms after VP.

department has recently been evaluated. The HEart failure and Audicor technology for Rapid Diagnosis and Initial Treatment (HEARD-IT) trial is an international multi-center study that has enrolled 995 patients presenting to the emergency room with acute dyspnea as their chief complaints [45]. This study showed that detection of S3 by acoustic cardiography was highly specific (specificity = 88.5%) for the diagnosis of acute heart failure adjudicated by cardiologists. However, the sensitivity of acoustic cardiography S3 for diagnosing clinical heart failure was relatively low (40%) in this study. Hence, although acoustic

cardiography S3 affected physicians' diagnostic confidence, it did not improve overall diagnostic accuracy for acute heart failure, largely because of its low sensitivity. However, acoustic cardiography may add independent and incremental value to the current diagnostic paradigm. In a secondary analysis of the HEARD-IT database [46], acoustic cardiography S3 significantly increased diagnosis sensitivity compared with auscultation in selected populations presenting with particular diagnostic challenges, namely, patients with intermediate BNP levels (100–499 pg/mL) and those who are obese (BMI N 30). When BNP levels

554

Y.-N. Wen et al. / International Journal of Cardiology 172 (2014) 548–560

Fig. 6. Acoustic cardiography report of episodes of snoring. (a) An acoustic cardiography report of snoring. (b) Time frequency analysis shows frequency of snoring very higher than heart sound.

fall into the “gray zone” (100–499 pg/mL), it is unreliable to be a sole mean to establish heart failure due to its low sensitivity and specificity [47,48]. Acoustic cardiogram supplements BNP diagnostic deficiency in indeterminate range. Prior investigators documented that when BNP was in an intermediate range, acoustic cardiogram can improve diagnostic sensitivity and specificity by 20% and 25%, respectively [29,46,47]. In 171 patients with either chronic or acute decompensated heart failure, a score derived from the products of 4 acoustic cardiography and ECG parameters (QRS duration × QR interval × S3 Strength × %EMAT) performed better than QRS duration alone with a 90% specificity and 77% sensitivity in detecting LV dysfunction [49]. In terms of prognosis, results from the HEARD-IT trial did not show any significant independent prognostic information provided by

acoustic cardiography S3. Nevertheless, we believe the 90-day followup period in this study may be too short to confer any definitive conclusion. As the prognostic significance of auscultation S3 has been demonstrated consistently in large clinical trials with longer follow-up, and acoustic cardiography is beyond doubt more accurate than physical auscultation in detecting S3, an equivalent or even superior robustness of acoustic cardiography in HF prognostication should be expected, although it remained to be proven by longer-term follow-up data. 4.1.3. Ambulatory heart failure monitoring Ambulatory acoustic cardiography (Audicor AM) is a new technique that enables continuous recording of the hemodynamic and acoustic parameters in conjunction with a 3-lead ECG. In asymptomatic subjects,

Y.-N. Wen et al. / International Journal of Cardiology 172 (2014) 548–560

555

Fig. 7. Heart rate and S4 trend during exercise tolerant test. (a) A patient without coronary artery disease by angiography. As shown, no diastolic heart sound was detected. (b) A patient with coronary artery disease by angiography. As shown, S4 was detected as heart rate increased.

S3 was significantly more prevalent in young patients (b40 years), especially during sleep, while nocturnal S4 was significantly more prevalent in those age N 40 years. In contrast, time intervals reflecting systolic function showed less circadian variation and less worsening with age [50]. In heart failure patients, diurnal variation of the systolic parameters of acoustic cardiography was diminished, probably related to a constant stimulation of, ambulatory acoustic information can aid in home monitoring of heart failure patients and prevention of clinical events by picking up early acoustic “signal” of heart failure decompensation to allow timely treatment warrants further investigations.

4.2. Detection of ischemia Given that acute impairment of LV wall motion and elevation of LV filling pressure are often earliest hemodynamic indicators of ischemia [52,53], a mechanistic link exists between diastolic heart sound and ischemic heart disease. Indeed, diastolic heart sounds occur more frequently in patients with angina, acute myocardial infarction, and during coronary angioplasty [54–57]. Akay et al. reported that diastolic heart sound detected by physical auscultation offered a diagnostic accuracy of 78% for coronary artery disease [58]. An advantage of acoustic

556

Y.-N. Wen et al. / International Journal of Cardiology 172 (2014) 548–560

Fig. 8. 12-lead acoustic cardiographic report. A. 10 second heart sound tracing suggesting presence of S3; B. ECG and acoustic measurements; C sound snapshot representative waveform of heart sound, in resynchronization with ECG; D An interpretative statements of both ECG and acoustic findings; E A 3D heart image illustrating location of myocardial infarction; F LVH scale showing strength of ECG evidence of LVH. If there is no evidence of LVH and MI, E and F will not be shown in the report.

cardiogram over physical auscultation is that it can quantitatively measure diastolic heart sound intensity over a period of time. Patients subjected to percutaneous coronary intervention-induced ischemia [58] developed new onset or increasing intensity of acoustic cardiography S3 or S4 that even preceded chest pain and ECG changes. Subsequent studies showed that acoustic cardiography may be useful as an adjunctive test to detect acute ischemia by improving the sensitivity (53% to 84%) and overall accuracy of ECG (76% to 81%) in patients with suspected acute coronary syndrome [59,60]. In addition to assisting ECG to improve the diagnosis of ischemia heart disease at rest, acoustic cardiography has also been used in stress testing. In a study of 59 patients with angiographically confirmed

coronary artery disease, diagnostic sensitivity of exercise treadmill ECG was merely 29%. Detection of acoustic cardiography S4 during exercise improved the diagnostic sensitivity of treadmill test to 53% without reducing specificity (92%). The combination of treadmill and S4 performed slightly better (sensitivity 68%) than S4 alone at the expense of lowering specificity (84%) [61]. Traditionally, echocardiography and radiocuclide are employed in conjunction with ETT to detect underlying CAD, due to the low sensitivity by ETT alone. Although Zuber et al. failed to show superiority of exercise acoustic cardiography over stress echocardiography and nuclear stress testing, the complete diagnostic performance with additional acoustic parameters such as EMAT was yet to be determined. If its diagnostic performance can be confirmed by more

Fig. 9. 24 h ambulatory acoustic cardiography datasheet. The datasheet shows that ambulatory acoustic cardiography records electrical, acoustic and respiratory data simultaneously over a period of 24 h. The first 4 rows show capture of arrhythmia, including bradycardia, tachycardia, ectopics, and pauses. The middle rows are respiratory data of snoring, apnea, and arousal, followed by acoustic parameters of S3, S4, EMAT and %EMAT. The last few rows give descriptions trending of heat rate, respiratory rate, activity and position.

Y.-N. Wen et al. / International Journal of Cardiology 172 (2014) 548–560 Table 1 Overview of AUDICOR parameters. Parameters

Definition

Electromechanical activation time(EMAT) %EMAT LVST (left ventricular systolic time) %LVST SDI (systolic dysfunction index)

Measured in ms from the Q wave to peak intensity of S1 The ratio of EMAT to the RR interval Measured in ms from S1 to S2 The ratio of LVST to RR interval QRS duration ∗ QR interval ∗ %EMAT ∗ S3 score The scale of S3 intensity from 0 to 10 The scale of S4 intensity from 0 to 10 Measured in ms from S2 to next S1 The ratio of LVDT to RR interval Measured in ms from S2 to next P wave The ratio of PADT to RR interval Measured in ms from P wave to S1

S3 strength S4 strength LVDT (left ventricular diastolic time) %LVDT PADT (pre-atrial diastolic time) %PADT AAFT (accelerated atrial filling time)

The left column is the list of AUDICOR parameters. The right side shows the definition accordingly.

studies, exercise acoustic cardiography has the advantage of being less expensive and less operator dependent than stress imaging modalities.

4.3. Optimization of cardiac resynchronization therapy Cardiac resynchronization therapy (CRT) is an effective therapy for patients with advanced heart failure refractory to pharmacological therapy with proven benefit in survival and symptomatic status [62]. An optimal AV delay setting is recognized as an important determinant of CRT response [63]. Conventionally, echocardiography is the most widely accepted method to program AV delay [64], but is limited by its relatively high cost. Most recently, acoustic cardiography has been used as an alternative technique to echocardiography to perform AV optimization, with its unique advantage of being faster, less operator dependent, while yielding results in CRT optimization comparable to echocardiography [65,66]. Furthermore, there is evidence suggesting that acoustic cardiography is superior to typical “out-of-the-box” settings to improve response to CRT by optimizing AV and VV delay [67]. Another application of acoustic cardiography in CRT and pacemaker therapy is the detection of phrenic nerve stimulation (PNS). It is reported

557

that PNS occurs frequently in patients implanted with pacemakers as a result of close location between phrenic nerve and pacing lead [68,69]. It has been documented that 24% of CRT patients have this problem which may be symptomatic or asymptomatic [68]. PNS not only cause symptoms such as muscle twitching, hiccups, and general malaise, but also adversely affects hemodynamics. Occasionally, PNS may be difficult to diagnose especially when symptoms are nonspecific or absent. Acoustic cardiograph objectively and accurately detects PNS. A prolonged EMAT is associated with the occurrence of PNS and impaired LV function [70].

4.4. Other applications Clinical utility of acoustic cardiography has been extended for the detection of other diseases including LV hypertrophy [71], constrictive pericarditis [72], sleep apnea [22,51] and ventricular fibrillation [73,74]. Prolongation of acoustic cardiography EMAT is strongly correlated with the presence of LVH [71]. It was reported that acoustic cardiography in conjunction with BNP outperformed ECG Cornell criteria, BNP or acoustic cardiography alone for LVH detection. Acoustic cardiography was shown to be sensitive to detect high-pitched pericardial knock in patients with constrictive pericarditis, with better performance than auscultation by cardiologist [72]. With newly developed ambulatory recording technique, acoustic cardiography was able to detect obstructive sleep apnea in hospital or home setting. Conventionally, an overnight sleep study observed by a well-trained technician with polysomnography is a standard method to detect sleeping-disordered breathing (SDB). Using wavelet techniques and time-frequency analysis of respiratory data, (Fig. 6), acoustic cardiography accurately captures respiratory events. As introduced above, ambulatory acoustic cardiography is a novel and simple tool to screen SDB, without need of overnight supervision. It can be even performed at home-setting. In a study involving 85 heart failure patients, ambulatory acoustic cardiography is comparable to polysomnography to diagnose SDB [22]. Using ambulatory acoustic cardiographic monitoring, sleep apnea is more prevalent in heart failure patients in heart failure than the normal subject [51], which is consistent with the findings by other studies. Presence of heart sounds during ventricular fibrillation is an indicator of successful cardiac conversion. Acoustic cardiography, as a

Fig. 10. Acoustic cardiography parameters in each cardiac cycle. EMAT—electromechanical activation time; PADT—pre-atrial diastolic time; LVDT—left ventricular diastolic time; LVST—left ventricular systolic time; AAFT—accelerated atrial filling time.

558

Y.-N. Wen et al. / International Journal of Cardiology 172 (2014) 548–560

Table 2 Cutoff values of 4 important AUDICOR parameters. Parameters EMAT

%EMAT

SDI

S3

S4

Snapshot 24 hour monitoring Snapshot 24 hour monitoring Snapshot 24 hour monitoring (%SDI N 5) Snapshot 24 hour monitoring (%S3 detected) Snapshot 24 hour monitoring (%S4 detected)

Day Night Day Night Day Night Day Night Day Night

Normal

Abnormal

b83 ms ± 17 b90 ms ± 10 b89 ms ± 11 b9.5% ± 2.6 b11.4% ± 1.8 b9.4% ± 1.6 b5 b9.2% ± 11.3 b16.9% ± 25.3 b5 b2.2% ± 2.2 b3.6% ± 5.3 b5 b9.2% ± 11.3 b16.9% ± 25.3

N120 ms N117 ms ± 25 N122 ms ± 27 N14% ± 3.8 N15.3% ± 3.0 N15.6% ± 4.5 ≧5 N42.5% ± 35.6 N49.1% ± 34.7 ≧5 N9.5% ± 12.6 N13.9% ± 21.7 ≧5 N7.7% ± 9.7 N11.8% ± 18.6

This table shows the cut-off values for EMAT, SDI, S3, S4 in both snapshot and ambulatory monitoring. For ambulatory monitoring, each parameter has different diurnal reference values.

sensitive heart sound auscultation tool, can objectively monitor heart sound during ventricular fibrillation conversion and evaluated resuscitation outcome [73,74].

5. Limitation Acoustic cardiography is still with several limitations. Firstly, data quality is influenced by endogenous and exogenous noises, including breathing and background speech. It is essential to have subject being quiet and still during recording. However, in patients with advanced heart failure, they are quite difficult to keep an idea examination position and relaxed, due to apnea and orthopnea. These factors consequently overestimate diastolic heart sounds. Then, despite that acoustic cardiography parameters are specific for heart failure, the sensitivity is still low (not exceed 55% by current studies), which is unsatisfactory as a diagnostic tool. Lastly, the sample sizes of current studies are not large enough to establish acute cut-off value for each parameter.

6. Future direction The prognostic role of acoustic cardiography parameters, especially EMTA, LVST and S3, is relatively unexplored. Further long-term followup data is required to define the role of acoustic cardiography as a cost-effective bedside tool to identify high-risk heart failure patients to supplement physical examination, which is often inaccurate. As of today, few studies have addressed the diagnostic accuracy and comparative cost-effectiveness of acoustic cardiography as an adjunctive tool to diagnosis and risk stratify patients with suspected ischemic heart disease during exercise stress testing. Ambulatory acoustic cardiography is the latest advance of the novel technique to allow ambulatory cardiac function monitoring. With its ease of use and relatively low cost, ambulatory acoustic cardiography can be used for heart failure follow-up in both hospital and home settings, particularly in detecting acute decompensation episodes at nights. 7. Conclusion Acoustic cardiography is newly technique which combines acoustic and electric information to detect and characterize heart sound. It correlates gold standard to identify heart failure and improves its management. Given that it is cost-effective and easy to use, it can serve as a screening tool to assist HF diagnosis, especially when echocardiography and other invasive investigations are unavailable. References

Fig. 11. Comparison of acoustic parameters in normal subject and heart failure patient. As shown, in heart failure patient, EMAT is prolonged, LVST is shorten, and S3 is present.

[1] Hanna IR, Silverman ME. A history of cardiac auscultation and some of its contributors. Am J Cardiol 2002;90(3):259–67. [2] Amin DSM, Fethi BR. Features for heartbeat sound signal normal and pathological. Recent Pat Comput Sci 2008;1:1–8. [3] Singh J, Anand RS. Computer aided analysis of phonocardiogram. J Med Eng Technol 2007;31:319–23. [4] Marcus GM, Vessey J, Jordan MV, et al. Relationship between accurate auscultation of a clinically useful third heart sound and level of experience. Arch Intern Med 2006;166(6):617–22. [5] Ishmail AA, Wing S, Ferguson J, Hutchinson TA, Magder S, Flegel KM. Interobserver agreement by auscultation in the presence of a third heart sound in patients with congestive heart failure. Chest 1987;91(6):870–3. [6] Avendano-Valencia LD, Ferrero JM, Castellanos-Dominguez G. Improved parametric estimation of time-frequency representations for cardiac murmur discrimination. Comput Cardiol 2008;35:157–60. [7] Balasubramaniam D, Nedumaran D. Efficient computation of phonocardiographic signal analysis in digital signal processor based system. Int J Comput Theory Eng 2010;2(4):1793–8201. [8] Emmanuel BS. A review of signal processing techniques for heart sound analysis in clinical diagnosis. J Med Eng Technol 2012;36(6):303–7. [9] Meziani F, Debbal SM, Atbi A. Analysis of phonocardiogram signals using wavelet transform. J Med Eng Technol 2012;36(6):283–302. [10] Debbal SM, Bereksi-Reguig F. Computerized heart sound analysis. Comput Biol Med 2008;38(2):263–80.

Y.-N. Wen et al. / International Journal of Cardiology 172 (2014) 548–560 [11] Lok CE, Morgan CD, Ranganathan N. The accuracy and interobserver agreement in detecting the ‘gallop sounds’ by cardiac auscultation. Chest 1998;114(5):1283–8. [12] Garrard CL, Weissler AM, Dodge HT. The relationship of alterations in systolic time intervals to ejection fraction in patients with cardiac disease. Circulation 1970;62: 455–62. [13] Lewis RP, Leighton RF, Forester WF, et al. Systolic time intervals. In: Weissler AM, editor. Non-invasive cardiology. New York: Grune and Stratton; 1974. p. 301–68. [14] Lewis RP, Boudoulas H, Welch TG, et al. Usefulness of systolic time intervals in coronary artery disease. Am J Cardiol 1976;37:787–96. [15] Lewis RP, Rittogers SE, Froester WF, et al. A critical review of the systolic time intervals. Circulation 1977;56:146–58. [16] Martin CE, Shaver JA, Thompson ME, et al. Direct correlation of external systolic time intervals with internal indices of ventricular function in man. Circulation 1971;44: 419–31. [17] McDonald IG, Hobson ER. A comparison of the relative value of non-invasive techniques—echocardiography, systolic time intervals, and apexcardiography—in the diagnosis of primary myocardial disease. Am Heart J 1974;88:454–62. [18] Stack RS, Lee CC, Reddy BP, et al. Left ventricular performance in coronary artery disease evaluated with systolic time intervals and echocardiography. Am J Cardiol 1976;37:331–9. [19] Stack RS, Sohn YH, Weissler AM. Accuracy of systolic time intervals in detecting abnormal left ventricular performance in coronary artery disease. Am J Cardiol 1981;47:603–9. [20] Weissler AM, Harris WS, Schoenfeld CD. Systolic time intervals in heart failure in man. Circulation 1968;37:149–59. [21] Weissler AM, Harris WS, Schoenfeld CD. Bedside technics for the evaluation of ventricular function in man. Am J Cardiol 1969;23:577–83. [22] Dillier R, Baumann M, Young M, et al. Continuous respiratory monitoring for sleep apnea screening by ambulatory hemodynamic monitor. World J Cardiol 2012;4(4): 121–7 [26]. [23] Efstratiadis S, Michaels AD. Computerized acoustic cardiographic electromechanical activation time correlates with invasive and echocardiographic parameters of left ventricular contractility. J Card Fail 2008;14(7):577–82. [24] Chao TF, Sung SH, Cheng HM, et al. Electromechanical activation time in the prediction of discharge outcomes in patients hospitalized with acute heart failure syndrome. Intern Med 2010;49(19):2031–7. [25] Shapiro M, Moyers B, Marcus GM, et al. Diagnostic characteristics of combining phonocardiographic third heart sound and systolic time intervals for the prediction of left ventricular dysfunction. J Card Fail 2007;13(1):18–24. [26] Collins SP, Lindsell CJ, Peacock WF, et al. The combined utility of an S3 heart sound and B-type natriuretic peptide levels in emergency department patients with dyspnea. J Card Fail 2006;12(4):286–92. [27] Shah SJ, Nakamura K, Marcus GM, et al. Association of the fourth heart sound with increased left ventricular end-diastolic stiffness. J Card Fail 2008;14(5):431–6. [28] Zuber M, Kipfer P, Jost CA. Systolic dysfunction: correlation of acoustic cardiography with Doppler echocardiography. Congest Heart Fail 2006;12(Suppl. 1):14–8. [29] Kosmicki DL, Collins SP, Kontos MC, et al. Noninvasive prediction of left ventricular systolic dysfunction in patients with clinically suspected heart failure suing acoustic cardiography. Congest Heart Fail 2010;16:249–53. [30] Collins SP, Lindsell CJ, Kontos MC, et al. Bedside prediction of increased filling pressure using acoustic electrocardiography. Am J Emerg Med 2009;27(4): 397–408. [31] Collins SP, Kontos MC, Michaels AD, et al. Utility of a bedside acoustic cardiographic model to predict elevated left ventricular filling pressure. Emerg Med J 2010; 27(9):677–82. [32] Wang S, Lam YY, Liu M, et al. Acoustic cardiography helps to identify heart failure and its phenotypes. Int J Cardiol 2013;167(3):681–6. [33] Toggweiler S, Odermatt Y, Brauchlin A, et al. The clinical value of echocardiography and acoustic cardiography to monitor patients undergoing anthracycline chemotherapy. Clin Cardiol 2013;36(4):201–6. [34] Shapiro M, Moyers B, Marcus GM, et al. Diagnostic characteristics of combing phonocardiographic 3rd heart sound and systolic time intervals for the prediction of LV dysfunction. J Card Fail 2007;13(1):18–24. [35] Shah SJ, Marcus GM, Gerber IL, et al. Physiology of the third heart sound: novel insights from tissue Doppler imaging and invasive LV hemodynamics. J Am Soc Echocardiogr 2008;21(4):394–400. [36] Marcus GM, Gerber IL, McKeown BH, et al. Association between phonocardiographic third and fourth heart sounds and objective measures of left ventricular function. JAMA 2005;293(18):2238–44. [37] Tribouilloy CM, Enriquez-Sarano M, Mohty D, et al. Pathophysiologic determinants of third heart sounds: a prospective clinical and Doppler echocardiographic study. Am J Med 2001;111(2):96–102. [38] Patel R, Bushnell DL, Sobotka PA. Implications of an audible third heart sound in evaluating cardiac function. West J Med 1993;158(6):606–9. [39] Drazner MH, Rame JE, Dries DL. Third heart sound and elevated jugular venous pressure as markers of the subsequent development of heart failure in patients with asymptomatic left ventricular dysfunction. Am J Med 2003;114(6):431–7. [40] Rame JE, Sheffield MA, Dries DL, et al. Outcomes after emergency department discharge with a primary diagnosis of heart failure. Am Heart J 2001;142(4):714–9. [41] Glover DR, Littler WA. Factors influencing survival and mode of death in severe chronic ischaemic cardiac failure. Br Heart J 1987;57(2):125–32. [42] Johnston M, Collins SP, Storrow AB. The third heart sound for diagnosis of acute heart failure. Curr Heart Fail Rep 2007;4(3):164–8.

559

[43] Michaels AD, Khan FU, Moyers B. Experienced clinicians improve detection of third and fourth heart sounds by viewing acoustic cardiography. Clin Cardiol 2010;33(3): 36–42. [44] Peacock WF, Harrison A, Moffa D. Clinical and economic benefits of using Audicor S3 detection for diagnosis and treatment of acute decompensated heart failure. CHF 2006;12:32–6. [45] Collins SP, Peacock WF, Lindsell CJ, et al. S3 detection as a diagnostic and prognostic aid in emergency department patients with acute dyspnea. Ann Emerg Med 2009; 53(6):748–57. [46] Maisel AS, Peacock WF, Shah KS, et al. Acoustic cardiography S3 detection use in problematic subgroups and B-type natriuretic peptide “gray zone”: secondary results from the Heart failure and Audicor technology for Rapid Diagnosis and Initial Treatment Multinational Investigation. Am J Emerg Med 2011;29(8):924–31. [47] Zuber M, Kipfer P, Attenhofer Jost CH. Usefulness of acoustic cardiography to resolve ambiguous values of B-type natriuretic peptide levels in patients with suspected heart failure. Am J Cardiol 2007;100(5):866–9. [48] Coste J, Jourdain P, Pouchot J. A gray zone assigned to inconclusive results of quantitative diagnostic tests: application to the use of brain natriuretic peptide for diagnosis of heart failure in acute dyspneic patients. Clin Chem 2006;52(12):2229–35. [49] Zuber M, Attenhofer Jost CH, Kipfer P, Collins SP, Michota F, Peacock WF. Acoustic cardiography augments prolonged QRS duration for detecting left ventricular dysfunction. Ann Noninvasive Electrocardiol 2007;12(4):316–28. [50] Dillier R, Zuber M, Arand P, Erne S, Erne P. Assessment of systolic and diastolic function in asymptomatic subjects using ambulatory monitoring with acoustic cardiography. Clin Cardiol 2011;34(6):384–8. [51] Dillier R, Zuber M, Arand P, Erne S, Erne P. Assessment of systolic and diastolic function in heart failure using ambulatory monitoring with acoustic cardiography. Ann Med 2011;43(5):403–11. [52] Barry WH, Brooker JZ, Alderman EL, Harrison DC. Changes in diastolic stiffness and tone of the left ventricle during angina pectoris. Circulation 1974;49:255–63. [53] Hauser AM, Gangadharan V, Ramos RG, Gordon S, Timmis GC, Dudlets P. Sequence of mechanical, electrocardiographic and clinical effects of repeated coronary artery occlusion in human beings: echocardiographic observations during coronary angioplasty. J Am Coll Cardiol 1985;5:193–7. [54] Gould L, Umali F, Gomprecht RF. The presystolic gallop in acute myocardial infarction. Angiology 1972;23:549–53. [55] Harris IS, Lee E, Yeghiazarians Y. Phonocardiographic timing of third and fourth heart sounds during acute myocardial infarction. J Electrocardiol 2006;39:305–9. [56] Karebsheh S, Michaels AD. Acoustic cardiographic indices of transmyocardial ischemia during percutaneous coronary intervention. Acute Card Care 2011; 13(1):3–8. [57] Akay M, Akay YM, Gauthier D, et al. Dynamics of diastolic sounds caused by partially occluded coronary arteries. IEEE Trans Biomed Eng 2009;56(2):513–7. [58] Lee E, Michaels AD, Selvester RH, Drew BJ. Frequency of diastolic third and fourth heart sounds with myocardial ischemia induced during percutaneous coronary intervention. J Electrocardiol 2009;42(1):39–45. [59] Lee E, Drew BJ, Selvester RH, Michaels AD. Sequence of electrocardiographic and acoustic cardiographic changes and angina during coronary occlusion and reperfusion in patients undergoing percutaneous coronary intervention. Ann Noninvasive Electrocardiol 2009;14(2):137–46. [60] Lee E, Drew BJ, Selvester RH, Michaels AD. Diastolic heart sounds as an adjunctive diagnostic tool with ST criteria for acute myocardial ischemia. Acute Card Care 2009; 11(4):229–35. [61] Zuber M, Erne P. Acoustic cardiography to improve detection of coronary artery disease with stress testing. World J Cardiol 2010;2(5):118–24 [26]. [62] Wells G, Parkash R, Healey JS, et al. Cardiac resynchronization therapy: a metaanalysis of randomized controlled trials. Can Med Assoc J 2011;183:421–9. [63] Auricchio A, Ding J, Spinelli JC, et al. Cardiac resynchronization therapy restores optimal atrioventricular mechanical timing in heart failure patients with ventricular conduction delay. J Am Coll Cardiol 2002;39(7):1163–9 [3]. [64] Waggoner AD, de las Fuentes L, Davila-Roman VG. Doppler echocardiographic methods for optimization of the atrioventricular delay during cardiac resynchronization therapy. Echocardiography 2008;25(9):1047–55. [65] Hasan A, Abraham WT, Quinn-Tate L, Brown L, Amkieh A. Optimization of cardiac resynchronization devices using acoustic cardiography: a comparison to echocardiography. Congest Heart Fail 2006;12(1):25–31. [66] Zuber M, Toggweiler S, Quinn-Tate L, Brown L, Amkieh A, Erne P. A comparison of acoustic cardiography and echocardiography for optimizing pacemaker settings in cardiac resynchronization therapy. Pacing Clin Electrophysiol 2008;31(7):802–11. [67] Toggweiler S, Zuber M, Kobza R, et al. Improved response to cardiac resynchronization therapy through optimization of atrioventricular and interventricular delays using acoustic cardiography: a pilot study. J Card Fail 2007;13(8):637–42. [68] Gurevitz O, Nof E, Carasso S, et al. Programmable multiple pacing configurations help to overcome high left ventricular pacing thresholds and avoid phrenic nerve stimulation. Pacing Clin Electrophysiol 2005;28(12):1255–9. [69] Sanchez-Quintana D, Cabrera JA, Climent V, et al. How close are the phrenic nerves to cardiac structures? Implications for cardiac interventionalists. J Cardiovasc Electrophysiol 2005;16(3):300–14. [70] Zuber M, Roos M, Kobza R, Toggweiler S, Meier R, Erne P. Detection and hemodynamic significance of cardiac pacemaker-induced phrenic nerve stimulation. Congest Heart Fail 2010;16(4):147–52.

560

Y.-N. Wen et al. / International Journal of Cardiology 172 (2014) 548–560

[71] Rogers RK, Collins SP, Kontos MC, Zuber M, Arand P, Michaels AD. Diagnosis and characterization of left ventricular hypertrophy by computerized acoustic cardiography, brain natriuretic peptide, and electrocardiography. J Electrocardiol 2008;41(6):518–25. [72] Michaels AD, Viswanathan MN, Jordan MV, Chatterjee K. Computerized acoustic cardiographic insights into the pericardial knock in constrictive pericarditis. Clin Cardiol 2007;30(9):450–8.

[73] Kobza R, Roos M, Toggweiler S, Zuber M, Erne P. Recorded heart sounds for identification of ventricular tachycardia. Resuscitation 2008;79(2):265–72. [74] Karabsheh S, Neuharth RM, Hamdan MH, Michaels AD. Acoustic cardiographic recording during ventricular fibrillation. Resuscitation 2008;78(3):374–7.

Beyond auscultation: acoustic cardiography in clinical practice.

Cardiac auscultation by stethoscope is widely used but limited by low sensitivity and accuracy. Phonocardiogram was developed in an attempt to provide...
3MB Sizes 0 Downloads 3 Views