Modification of Human Left Ventricular Relaxation by Small-Amplitude, Phase-Controlled Mechanical Vibration on the Chest Wall
Background Direct clinical manipulation to improve an impairment of left ventricular (LV) relaxation has not been reported. We investigated whether the LV relaxation rate in humans could be modulated by phase-controlled mechanical vibration applied to the patient's anterior chest wall and whether there are some quantitative differences in the responses of normal (N), hypertrophied (H), and failing (F) ventricle.
Methods and Results In 46 patients (N, 10; H, 18 [hypertrophic cardiomyopathy]; F, 18 [heart failure]), the vibrator was attached to the precordium and a 50-Hz, 2-mm sinusoidal mechanical vibration was applied, with the timing restricted from the onset of isovolumic relaxation to end-diastole during cardiac catheterization. Heart rate and peak LV pressure showed no difference with vibration. However, in all patients, precordial vibration caused an acceleration of the LV pressure fall. The magnitude of the induced reduction of the time constant of LV pressure decay (ΔT) was larger (P<.01) in H and F than in N (4.6±2.3, 4.0±1.6, and 0.6±1.5 ms for H, F, and N, respectively). ΔT correlated strongly with the magnitude of impaired relaxation and the magnitude of transmitted vibration to the ventricle.
Conclusions Phase-controlled, small-amplitude vibration on the chest wall can directly modulate LV relaxation rate, especially in those with hypertrophy or failing ventricle.
In many cardiac diseases, impaired relaxation has been reported to be one of the major determinants in the pathophysiology of heart failure, reducing the coronary perfusion and reducing the mitral inflow to the ventricle.1 2
Recently, in experimental studies, we have reported that vibration from the onset of isovolumic relaxation to end diastole (phase-controlled vibration) caused an acceleration of the ventricular relaxation rate, especially at serious levels of impaired relaxation under globally ischemic3 or in propranolol-injected, acute failing heart preparations.4 Moreover, in the preparation of regional ischemia, the phase-controlled vibration induced an instantaneous increase in coronary blood flow that depended on the input mechanical energy and the severity of the ischemia.5 From the point of view of mechanoenergetics, vibration to the failing heart has been demonstrated to be unique in improving the oxygen cost of contractility by decreasing the pressure-volume area–related myocardial oxygen consumption in an experimental study using isolated canine ventricles (see discussion on the clinical feasibility and clinical implication).6 7
The induced response of the ventricle observed in these experiments was, to our surprise, remarkable even when the input amplitude of the external vibration was set as small as that of the first heart sound. However, it has never been determined whether we can actually modify the LV relaxation in patients by extrapolating these experimental results to the clinical situation.
In this study, we examined whether we could input phase-controlled vibration into the human ventricle by applying the mechanical vibrator to the anterior chest wall and whether we could modulate the LV pressure of the patients, especially in those with hypertrophy and failing ventricle.
Clinical studies were carried out on 46 patients. They were divided into three groups: (1) those who had chest pain but were found to be normal through examination during hospitalization (N group; n=10; female/male, 0/10); (2) 18 patients with HCM (H; female/male, 3/15) diagnosed by echocardiography, MRI, and routine cardiac catheterization, including left or right myocardial biopsy; and (3) 18 patients with depressed LV function (F; female/male, 10/8), ie, an ejection fraction by levography of <50%, LV end-systolic volume index >50 mL/m2 of body surface area, and a history of experiencing the limited daily activity of NYHA functional classes II and III. NYHA functional class at the most serious condition in the past (NYHAmax) was higher in F (mean±SD, 3.4±0.5; P<.01) than in N (1.5±0.5) or H (1.9±0.7). This group included 7 patients with dilated cardiomyopathy, 3 with aortic regurgitation of grade 3 by Sellars and associates,8 1 with mitral regurgitation grade 3, 5 with coronary artery stenosis, and 2 with old myocardial infarction. The basal hemodynamics of patients as measured during cardiac catheterization are summarized in Table 1⇓. All patients were in sinus rhythm. LV ejection fraction, determined by the method of Kennedy et al9 from ventriculography in the 30° right anterior oblique projection, ranged from 23% to 86% (mean±SD, 60.0±17.8%). All patients gave informed consent to the study, which had been approved by the Institutional Committee on Ethics in Human Research.
All medications were withheld for at least 12 hours before the hemodynamic study. After pressure measurements in the routine cardiac catheterization and before cineangiography, the effect of the precordial phase-controlled vibration on LV relaxation was examined (perturbation study in this report). An intraesophageal miniature acceleration sensor (Emic 540M, Shin-Nippon Sokki) was optimally positioned for the detection of LV posterobasal wall vibration, usually at 30 to 37.5 cm from the incisor teeth. An intraventricular catheter-tipped micromanometer (PC 370, Millar Instruments) was also introduced into the left ventricle. The respiration curve was obtained by a nasal thermistor (type 45257, NEC San-ei Instruments Co Ltd).
At the beginning of the protocol, the transmission of the precordially applied mechanical vibration (50 Hz, 2-mm amplitude) to the ventricle was confirmed. That is, the vibrator on the patient's precordium (third or fourth intercostal space of the left parasternal border) was driven for brief periods (40 to 60 ms) at different times throughout the cardiac cycle. We checked whether such precordial mechanical perturbations induced either LVPosc and/or posterobasal wall vibration as detected by intraesophageal vibration sensor. Then we applied 50-Hz, 2-mm-amplitude, sinusoidal, phase-controlled mechanical oscillation to the same input site of the precordium, measuring pressure data under the quiescent (previbration) condition and under 15 seconds of vibration during a voluntary holding of the breath (≈25 seconds) near end expiration. The magnitude of expiration during and before vibration was held constant by monitoring of the respiration curve at the bedside oscilloscope. The timing of the vibration was carefully adjusted throughout the study as described below.
For this clinical study, we made a special mechanical vibrator system. The arm of the supporting part for the vibrator and the vibrator itself can be rotated three-dimensionally to ensure perpendicular attachment to the patient's precordium. The tip of the vibrator was made of silicone rubber, which had been molded from five patients of different body sizes beforehand. We usually selected the one (2×4×2 cm, slightly concave configuration at the attached surface) that best fit the intercostal space of the patient receiving the application. The magnitude of the precordially applied vibration was monitored by a miniature vibration sensor (Emic 540, Shin Nippon Sokki Co Ltd) at the shaft of the vibrator. The vibrator was driven by an intermittent sine-wave signal, which was produced by a signal passed through an on/off gate that in turn was triggered by the R wave of the ECG. The timing of the vibration was restricted to the period from the beginning of isovolumic relaxation to the beginning of the subsequent systole. We defined the beginning of isovolumic relaxation to be the point at which LVP equals the value of the nadir at aortic pressure incisura. The onset of the subsequent systole was identified by the upstroke of LVP. We were careful not to apply vibration after this point. The timing of these starting and ending points was adjusted manually by monitoring of the superposed, calibrated waveforms of LV and aortic pressure and the precordial vibration on an oscillograph (2G66, NEC-Sanei Instruments Co Ltd).
All data obtained in this study were recorded on an analog tape recorder (type FE-3907W, Sony Magnescale Co Ltd) and, after low-pass filtering (50-Hz cutoff), on a thermal recorder (8M15, NEC San-ei Instruments Co Ltd) for later analysis. ECG, LVP, input vibration signal from the miniature vibration sensor (Shin Nippon Sokki, Emic 540M) at the shaft of the vibrator, and the induced vibration of the intraesophageal miniature vibration sensor were digitized at 1 kHz with a computer system (7T-17, NEC San-ei Instruments Co Ltd). There was no significant change in ventricular hemodynamics during vibration. The data from pressure measurement were averages of three to five beats from before and during vibration. We obtained the values described below from the digitized data.
Time constant of LVP fall (T)
We adopted the method of Weiss et al10 for the calculation of the time constant of LVP fall as an index of myocardial relaxation. With a signal processor (7T17, NEC San-ei Instruments Co Ltd), LVP was digitized at 1 kHz and fitted to an exponential decay by the least-squares method as described by Weiss et al:P(t)|<|=|>|P_|<|0|>||<|\times|>|\mathit|<|e|>|^|<|(|<|-|>|t/T)|>|where P(t) is instantaneous ventricular pressure at time t and P0 and T are constants. Time 0 is taken at the peak negative rate of LVP change (peak negative dP/dt), and data were taken until the pressure decreased to a level 5 mm Hg above the LVEDP.
Magnitude of the vibration-induced LVPosc at diastole
The vibration-induced response of the LVP decay has been reported to be dependent on the input amplitude of the mechanical vibration as well as on the underlying ventricular function.3 4 We estimated the effective input amplitude of the mechanical vibration to the left ventricle by measuring the superposed oscillation on LVP by a catheter-tipped micromanometer at middiastole.
Index from LV volume measurement
The LV volume was determined by the method of Kennedy and associates.9 LV end-diastolic volume index is the normalized value (per body surface area) of LV end-diastolic volume.
The wall thickness was measured at the crossing point of the perpendicular line at two thirds of the LV long axis in the right anterior oblique projection.
All values are expressed as mean±SD unless otherwise indicated. Multiple regression analysis and selection of variables were used to search for links between the magnitude of vibration-induced ΔT and the variables of each category listed in Table 1⇑. First, the parameter was searched by stepwise regression analysis (forward increment method) using age, NYHA functional class, and all basal hemodynamics in each category listed in Table 1⇑ (F-to-enter value, >4.00). Then multiple regression analysis was done, including the variables selected by the stepwise regression analysis. ANOVA was used to assess the difference in the hemodynamic variables among the groups. ANOVA was also used to assess the difference of the vibration-induced change of the variables among the groups. When statistically significant results were found, individual comparisons were made with Bonferroni's test. A value of P<.05 was considered statistically significant.
Fig 1⇓ shows the transmission of the precordial vibration to the ventricle. When the phase-controlled vibration (from the onset of isovolumic relaxation to end diastole) was applied, the intraesophageal vibration and the ventricular pressure (recorded at high gain) showed oscillations of identical frequencies (50 Hz) and equivalent durations. The delay of the transmission from precordial input to the output signals by intraesophageal and intraventricular sensors was <4 ms.
No arrhythmias were induced by precordial vibration, and no significant difference in heart rate was observed between before vibration and during the vibration condition. The precordial vibration evoked oscillations of LVP that ranged from 0.4 to 1.2 mm Hg.
Table 2⇓ shows the value of LVPosc, the change in heart rate, peak systolic LVP, LVEDP, minimum LVP, dP/dt, and the ventricular relaxation in applying the diastolic vibration. LVPosc was larger in the H group than in the F and N groups. There was no significant difference in induced change of heart rate, LVP, LVEDP, minimum LVP, or dP/dt between the groups. The magnitudes of the induced change of the relaxation time constant were larger in H and F than in the normal group (P<.01). There was no significant difference in pressure-related hemodynamics between previbration and postvibration conditions.
Fig 2⇓ shows a induced response of the LVP by application of diastolic vibration. In this patient, a 24-year-old woman with DCM showing alternating pulse (mechanical alternance), the vibration resulted in a decrease in diastolic pressure and an increase in systolic pressure as well as an improvement of the ventricular relaxation. However, this response in the systolic pressure was observed in only two patients.
Fig 3⇓ shows a superposition of the ventricular pressure before and during vibration in a 56-year-old man with HCM. The precordially applied vibration induced an acceleration of the ventricular relaxation rate (ΔT=4.3 ms) and a slight lowering of the ventricular minimal pressure. However, in this patient, there was no significant increase in systolic pressure. The T showed an instantaneous decrease when the vibration was applied, ie, the vibration caused an acceleration of the LVP fall.
In Fig 4⇓ (top), ΔTs in each group are shown. ΔTs in patients with HCM (H) and those with depressed function (F) are larger than those in patients with normal function (N, P<.01). The relationship between ΔT and T (ΔT=1.444−0.1×T, r=.708, P<.0001) is shown in Fig 4⇓, bottom.
From the multiple regression analysis using all variables listed in Table 1⇑, T, LVPosc, LVEDP, mean pulmonary arterial pressure (PAPmean), NYHA functional class, and LV dP/dt among these variables were shown to correlate significantly with the magnitude of the induced ΔT as given by the fitted equation|<|\Delta|>|T|<|=|>|(|<|-|>|0.104)T|<|+|>|(|<|-|>|4.072)LVP_|<|osc|>||<|+|>|(0.165)LVEDP|<|+|>|(|<|-|>|0.186)PAP_|<|mean|>||<|+|>|(|<|-|>|1.300)NYHA Class|<|+|>|(|<|-|>|0.001)LV dP/dt|<|+|>|9.514(r2=.837), where times are expressed in milliseconds and pressures in mm Hg. The coefficients from the multiple regression analysis are shown in Table 3⇓. Other variables were not related to ΔT. Here, the induced change in T has been shown to be strongly correlated with T at previbration condition (see Fig 4⇑, bottom) and LVPosc. The multiple regression equation in category H or F was equivalent to the equation for the entire category. Moreover, to confirm whether other factors not listed in Table 1⇑ could result in differences of ΔT in each group, the multiple regression analysis was repeated, including the diagnosis as a dummy parameter. The results were essentially the same as those without the dummy parameter. Therefore, the larger ΔT in H and F could be explained for the most part by the difference of LVPosc, T, and the net effect of the baseline features listed in Table 1⇑.
The reason that we adopted 50-Hz vibration in this study is based on the following observations obtained from the preliminary experimental study. In a propranolol-injected (0.05 mg/kg body wt IV) acute heart failure model of canine open-chest preparations (n=2), as used in a previous report,4 we examined the effects of the amplitude and frequency of phase-controlled vibration on the ventricular relaxation time constant. When we changed the amplitude while keeping the frequency at 50 Hz, T decreased as the amplitude increased, even though the slope seemed to possess some limitations in this decrement of time constant (Fig 5⇓, top). As for the relation between the applied frequency and the resultant change of the time constant, the decrease in T was maximal when the frequency was set at ≈50 Hz (Fig 5⇓, bottom; here the amplitude of the vibration was set as a constant at 2 mm). This optimal value of frequency was close to the resonant frequency during the relaxation period, although the value is changing instantaneously during the cardiac cycle.11 We extrapolated these results to the clinical study, setting 50 Hz as the input frequency. This phase-controlled mechanical vibration to the precordium did not induce chest pain at the attached portion, arrhythmias, or emotional excitement (no change in heart rate during the vibration).
Transmission of the Mechanical Vibration From Human Precordium to the Left Ventricle
This study is the first report demonstrating that phase-controlled mechanical vibration from the chest wall can modulate the relaxation of human left ventricle. Strictly speaking, the magnitude of the transmission of the precordially applied mechanical vibration to the human ventricle is strongly influenced by the physical properties of the interposing tissues between input site and ventricle. At present, we could not find any quantitative report on the transmission characteristics of the human thorax. Therefore, we evaluated the magnitude of an effective mechanical vibration on the ventricle by LVPosc, ie, superposed oscillations on the pressure signal measured by an intraventricular catheter-tipped micromanometer. From this study, we stress that the phase-controlled mechanical vibration on the chest wall is well transmitted to the ventricle and effectively vibrates the ventricular myocardium. The magnitude of LVPosc was larger in HCM patients than in those of normal and failing ventricle. We speculated that one of the reasons was the higher transmissibility of the mechanical vibration in the ventricular wall of those patients. Both an increase in the elasticity per unit myocardium12 and an increase in the thickness of the ventricular wall would result in larger transmissibility of the vibration from the precordium to the intraventricular lumen in HCM patients.
Speculated Mechanisms of Modulation of the Ventricular Relaxation
The reduced rate of the LV relaxation and/or the decreased LV compliance have been demonstrated in patients with heart failure from valvular disease, ischemic heart disease, hypertensive heart disease, and idiopathic cardiomyopathy.1 13 14 15 16 The impairment of the relaxation rate and ventricular filling occur early in the failing condition and can be identified before the onset of systolic dysfunction. Furthermore, a large portion of patients with clinical symptoms of heart failure have been reported to possess not only systolic impairment but also an impairment of relaxation and diastolic filling.17 18
In previous studies of the isolated ventricle with global ischemia, the phase-controlled vibration induced shortening of T and, under severe ischemia, a decrease in the chamber stiffness.3 Moreover, in an in situ open-chest canine preparation of a propranolol-injected acute heart failure model, the vibration caused an immediate improvement of the impaired relaxation and, in cases of severe heart failure, an immediate augmentation of systolic ventricular function.4
The clinical study reported here verifies that the magnitude of improvement in ventricular relaxation is larger in patients with hypertrophy or with depressed ventricular function. From the multiple regression analysis, the induced change in T has been shown to depend strongly on the severity of the baseline relaxation impairment and the amplitude of the vibration applied to the ventricle. These results are qualitatively consistent with the findings in the previous experimental report. Because no significant difference was observed between with and without the dummy parameter for each group in the regression analysis, the observed larger response of the ventricular relaxation rate in H and F seemed to be explained, in large part, by the inherent severity of the impaired relaxation.
It has been postulated that the ventricular relaxation is determined by triple control mechanisms: reduced calcium ion sequestration by the SR leading to prolonged activation of the force-generating cross-bridges, insensitivity to loading conditions, and spatial or temporal nonuniformity.2 19 All of these factors are likely to occur concurrently in human heart disease.1 20 Brutsaert et al21 reported that an abrupt increase in load resulted in more rapid relaxation in the isolated papillary muscle. Such load-dependent relaxation has also been demonstrated in intact canine heart.22 23 The load-dependent relaxation has been considered to relate to the function of the SR and load-induced cross-bridge backrotation from the strongly binding, force-producing state to the weakly binding, non–force-producing state.2 Several studies recognized the interplay of the cross-bridge kinetics as a dominant factor for the induced change in ventricular relaxation. Through a theoretical extension of the model on the cross-bridge kinetics by Gray, Gonda, and Cheung,24 25 26 Honda et al27 demonstrated that mechanical perturbation decreases the number of the force-producing, strongly binding cross-bridges of the myocardium in an amplitude-dependent manner. Moreover, Huxley and Simmonds28 reported that a quick change of the skeletal muscle length reduces the number of active cross-bridges (deactivation). This phenomenon has also been observed in a papillary muscle29 and the isolated ventricle.30 In a mechanoenergetic study using isolated canine left ventricle, Nishioka et al6 clearly demonstrated that the vibration affected cross-bridge kinetics but not the component of calcium flux, excitation-contraction coupling, or basal metabolism.
As for the modification of sarcoplasmic function when exposed to external vibration, which is another possible mechanism for the vibration-induced change of the ventricular relaxation rate, we regard it as less probable at present because in the previous experiment, the magnitude of the vibration-induced decrease in T became larger at severe myocardial ischemia.3 If the change of function of the SR played a dominant role in this perturbation study, then the aggravation and deterioration of SR function under severe myocardial ischemia would result in a reduced magnitude in the response to the sudden change of the mechanical load, ie, a smaller change of the relaxation rate when the mechanical vibration is applied. This is in sharp contrast to the observations in the previous study under global ischemia. Therefore, we speculate that the effects of the vibration are mainly through the promotion of the deactivation of active, force-producing cross-bridges by induced cyclic changes of the myocardium, although we cannot exclude the possibility of the functional interplay of the SR function under perturbation in these clinical situations.
Clinical Feasibility and Implications
Before we can draw any conclusions about the clinical value, it would be necessary to clarify the effects on oxygen consumption of changing the cross-bridge kinetics by external vibration. On this issue, Nishioka et al studied the effect of continuous6 and phase-controlled7 vibration, as used in this investigation, on ventricular function and myocardial oxygen consumption, measuring Emax and ventricular pressure-volume area. They concluded from these studies that the only interventions that improved the oxygen cost of the contractility (LV contractility per unit oxygen consumption) were hypothermia and mechanical external vibration, such as this perturbation method. In contrast, other inotropic drugs examined (catecholamines, digitalis, phosphodiesterase inhibitor,6 sodium channel opener31 ) failed to improve the oxygen cost of the contractility (personal communication, H. Suga, MD, 1993). Thus, we considered that the vibration rendered the cross-bridge cycling more efficient, and this intervention induced either functional or metabolical improvement of the ventricle.
In cases of severe heart failure, a vicious circle acts as an underlying mechanism of aggravating the function. The impaired relaxation in each beat plays an important role in this vicious circle, increasing the oxygen demand in every beat because of the higher diastolic pressure by delayed cross-bridge inactivation, reducing coronary perfusion and shifting the operating point to higher end-diastolic volume and/or decreasing diastolic ventricular compliance. Therefore, if the mechanical vibration can break this vicious circle in severe heart failure through the normalization of the impaired relaxation, then a functional improvement might be realistic. In experimental studies, we have observed beneficial effects of phase-controlled vibration,3 4 including the instantaneous improvement of coronary perfusion.5 However, we applied only brief periods (6 to 10 seconds) of mechanical perturbation in this study to examine the effect on the LVP decay. Therefore, further studies on this possibility using a much longer perturbation period are necessary.
In previous experiments, we have confirmed that the magnitude of the response induced by external vibration was related to the input energy of the vibration as well as to the functional severity of the ventricular failure.3 4 We should not take the magnitude of the ventricular response in this report as the maximum of the perturbation method. We did not study the response in patients with severe heart failure of NYHA functional class IV (the value of T in patients with severe heart failure has been reported to reach larger values of ≈80 ms32 ); moreover, the transmitted amplitude of the vibration to the ventricle in this study was rather small. Considering the input amplitude dependency reported here (see Equation 2), the induced modulation of the ventricle would become much more obvious at larger amplitudes of vibration. This possibility, however, needs further investigation.
In Equation 2, the estimate of LVEDP was positive, as if the elevation induced a decreased response in ΔT. However, the relation of ΔT versus LVEDP by univariate analysis was ΔT=−1.221+(−0.177)LVEDP (r=.473, P<.001). The result from the univariate analysis is consistent with those in previous reports,3 4 and we considered that the positive estimate of LVEDP would be from the complex interrelation between variables (especially with T) in the multiple regression analysis. From the clinica1 point of view, we considered that we should include these variables together in this analysis because of their pathophysiological importance.
1. We can apply phase-controlled mechanical vibration to the human left ventricle by precordial application of a vibrator without any objective or subjective side effects.
2. The LV relaxation rate can be modulated in the clinical setting by precordially applied small-amplitude, phase-controlled vibration.
3. The quantitative relationship between the induced response in human ventricular relaxation and other parameters was estimated by use of multiple regression analysis.
4. The improvement in relaxation rate is more prominent in patients with serious impairment of relaxation, ie, in those with hypertrophied ventricle or with depressed ventricular function.
Selected Abbreviations and Acronyms
|ΔT||=||change of T|
|LVEDP||=||LV end-diastolic pressure|
|T||=||relaxation time constant of LVP decay|
This study was supported by the Research Development Corporation of Japan, Technology Science Agency.
Presented in part at the 65th Scientific Sessions of the American Heart Association, New Orleans, La, November 16-29, 1992.
- Received February 29, 1996.
- Revision received August 16, 1996.
- Accepted August 24, 1996.
- Copyright © 1997 by American Heart Association
Grossman W. Diastolic dysfunction and congestive heart failure. Circulation. 1990;81(suppl III):III-1-III-7.
Brutsaert DL, Rademakers FE, Sys SU. Triple control of relaxation: implications in cardiac disease. Circulation.. 1984;69:190-196.
Kikuchi J, Koiwa Y, Takagi T, Honda H, Hoshi N, Butler JP, Takishima T. Effects of mechanical vibration on left ventricular diastolic properties during global ischemia. Am J Physiol.. 1992;263:H88-H95.
Takagi T, Koiwa Y, Kikuchi J, Honda H, Hoshi N, Butler JP, Takishima T. Diastolic vibration improves systolic function in cases of incomplete relaxation. Circulation.. 1992;86:1955-1964.
Koiwa Y, Honda H, Hoshi N, Naya T, Kamada E. The effect of diastolic vibration on the coronary flow rate in the canine heart with regional ischemia. J Cardiovasc Diagn Procedures.. 1994;12:10. Abstract.
Nishioka T, Goto Y, Hata K, Takasago T, Saeki A, Taylor TW, Suga H. Mechanoenergetics of negative inotropism of ventricular wall vibration in dog heart. Am J Physiol.. 1996;260:H583-H593.
Nishioka T, Goto Y, Hata K, Takasago T, Saeki A, Suga H. Improvement of the ventricular mechanical efficiency by diastolic vibration [in Japanese]. Jpn Circ J.. 1993;57:349. Abstract.
Sellers RD, Levy MJ, Amplatz K, Lillehei CW. Left retrograde cardiography in acquired cardiac disease: technics, indications and interpretation in 700 cases. Am J Physiol.. 1964;14:437-447.
Weiss JL, Frederiksen JW, Weisfeldt ML. Hemodynamic determinants of time course of fall in canine left ventricular pressure. J Clin Invest.. 1976;58:751-760.
Koiwa Y, Hashiguchi R, Ohyama T, Isoyama S, Satoh S, Suzuki H, Takishima T. Measurement of instantaneous viscoelastic properties by impedance frequency curve of the ventricle. Am J Physiol.. 1986;259:H672-H684.
Honda H, Koiwa Y, Takagi T, Kikuchi J, Hoshi N, Takishima T, Butler JP. Noninvasive measurement of left ventricular myocardial elasticity. Am J Physiol.. 1994;266:H881-H890.
Mann T, Goldberg S, Mudge GH Jr, Grossman W. Factors contributing to altered left ventricular diastolic properties during angina pectoris. Circulation.. 1979;59:1420. Abstract.
Carroll JD, Lang RM, Neumann AL, Bonow KM, Rajfer SI. The differential effects of positive inotropic and vasodilator therapy on diastolic properties in patients with congestive cardiomyopathy. Circulation.. 1985;74:815-825.
Paulus WJ, Goethals MA, Sys SU. Failure of myocardial inactivation: a clinical assessment in the hypertrophied heart. Basic Res Cardiol. 1992;82(suppl 2):145-161.
Lew WYW, Rasmussen CM. Influence of non-uniformity on rate of left ventricular pressure fall in the dog. Am J Physiol.. 1989;256:H222-H232.
Brutsaert DL, Sys SU. Relaxation and diastole of the heart. Physiol Rev.. 1989;69:1228-1315.
Ariel Y, Gaasch WH, Bogen DK, McMahon TA. Load-dependent relaxation with late systolic volume steps: servo-pump studies in the intact canine heart. Circulation.. 1987;75:1287-1294.
Zile MR, Gaasch WH. Load-dependent relaxation in conscious dogs. Am J Physiol.. 1991;261:H691-H699.
Cheung AS, Gray BF. The sliding filament model of muscle contraction, III: stability analysis and sinusoidal perturbations. J Theor Biol.. 1981;91:495-506.
Huxley AF, Simmonds RM. Proposed mechanism of force generation in striated muscle. Nature.. 1970;233:533-538.
Bodem R, Sonnenblick EH. Deactivation of contraction by quick releases in the isolated papillary muscle of the cat: effects of lever damping, caffeine and tetanization. Circ Res.. 1974;34:214-225.
Hunter WC, Janicki JS, Weber KT, Noordergraaf A. Systolic mechanical properties of the left ventricle: effects of volume and contractile state. Circ Res.. 1983;52:319-327.
De Tombe PP, Burkoff D, Hunter WC. Effects of calcium and EMD-53998 on oxygen consumption in isolated canine hearts. Circulation.. 1980;86:1945-1954.
Herrmann HC, Ruddy TD, Dec GW, Strauss HW, Boucher CA, Fifer MA. Diastolic function in patients with severe heart failure: comparison of the effects of enoximone and nitroprusside. Circulation.. 1987;75:1214-1221.