The Effect of Flow Acceleration on the Cyclic Variation of Blood Echogenicity Under Pulsatile Flow

Ultrasound in Med. & Biol., Vol. 39, No. 4, pp. 670–680, 2013 Copyright Ó 2013 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 0301-5629/$ - see front matter

http://dx.doi.org/10.1016/j.ultrasmedbio.2012.10.018

d

Original Contribution THE EFFECT OF FLOW ACCELERATION ON THE CYCLIC VARIATION OF BLOOD ECHOGENICITY UNDER PULSATILE FLOW CHIH-CHUNG HUANG,* CHEN-CHIH LIAO,* PO-YANG LEE,*y and CHO-CHIANG SHIH*y * Department of Electrical Engineering, Fu Jen Catholic University, New Taipei City, Taiwan; and y Graduate Institute of Applied Science and Engineering, Fu Jen Catholic University, New Taipei City, Taiwan (Received 26 June 2012; revised 22 September 2012; in final form 28 October 2012)

Abstract—It has been shown that the echogenicity of blood varies during a flow cycle under pulsatile flow both in vitro and in vivo. In general, the echogenicity of flowing whole blood increases during the early systole phase and then reduces to a minimum at late diastole. While it has been postulated that this cyclic variation is associated with the dynamics of erythrocyte aggregation, the mechanisms underlying this increasing echogenicity with flow velocity remain uncertain. The effect of flow acceleration has also been proposed as an explanation for this phenomenon, but no specific experiments have been conducted to test this hypothesis. In addition, the influence of ultrasonic attenuation on the cyclic variation of echogenicity requires clarification. In the present study, a Couette flow system was designed to simulate blood flowing with different acceleration patterns, and the flow velocity, attenuation, and backscattering coefficient were measured synchronously from 20%- and 40%-hematocrit porcine whole blood and erythrocyte suspensions using 35-MHz ultrasound transducers. The results showed ultrasonic attenuation exerted only minor effects on the echogenicity of blood under pulsatile flow conditions. Cyclic variations of echogenicity were clearly observed for whole blood with a hematocrit of 40%, but no variations were apparent for erythrocyte suspensions. The echogenicity did not appear to be enhanced when instantaneous acceleration was applied to flowing blood in any case. These findings show that flow acceleration does not promote erythrocyte aggregation, even when a higher peak velocity is applied to the blood. Comparison of the results obtained with different accelerations revealed that the cyclic variation in echogenicity observed during pulsatile blood flow may be jointly attributable to the effect of shear rate and the distribution of erythrocyte on aggregation. (E-mail: [email protected]) Ó 2013 World Federation for Ultrasound in Medicine & Biology. Key Words: Erythrocyte aggregation, Pulsatile flow, Cyclic variation, Ultrasonic attenuation, Echogenicity.

cations. Both the in vitro and in vivo properties of erythrocyte aggregation under steady flow conditions have been widely studied using ultrasound techniques because of their real-time and noninvasive capabilities (Shung and Thieme 1993). Because the ultrasonic backscatter signal is strongly related to the size of the rouleaux, the level of erythrocyte aggregation can be evaluated by measuring the echogenicity of blood. This has led to many studies demonstrating that erythrocyte aggregation in steadily flowing blood is dependent mainly on the hematocrit, turbulence, shear rate, plasma fibrinogen concentration, vessel-wall compliance and flow disturbance (Cloutier et al. 1996; Huang et al. 2009; Huang and Wang 2007; Yuan and Shung 1988a, 1988b). Some studies have also characterized erythrocyte aggregation under pulsatile flow conditions using ultrasound. However, the relationship between erythrocyte aggregation and ultrasonic backscatter signals from pulsatile flowing blood is not fully understood.

INTRODUCTION Erythrocyte aggregation in flowing blood is a reversible physiologic phenomenon. Many studies have indicated that abnormal erythrocyte aggregation can occur in some pathologies and diseases, such as vascular thrombosis and cardiovascular diseases (Chabanel et al. 1994; Demiroglu et al. 1996; Hahn et al. 1989), hypercholesterolemia (Bosch et al. 2001), diabetes (Cloutier et al. 2008; Le Devehat et al. 1990; Schmid-Sch€ onbein and Volger 1976), hypertension (Razavian et al. 1992), hyperlipidemia (Cloutier et al. 1997; Razavian et al. 1994), morbid obesity (Samocha-Bonet et al. 2004) and heavy smoking (Li et al. 2011). This makes it crucial to characterize erythrocyte aggregation in flowing blood in clinical appliAddress correspondence to: Chih-Chung Huang, Department of Electrical Engineering, Fu Jen Catholic University, 510 Chung Cheng Rd, Hsin Chuang District, New Taipei City 24205, Taiwan. E-mail: [email protected] 670

Flow acceleration on blood aggregation d C.-C. HUANG et al.

Under steady flow, the effect of shear rate has been suggested as a crucial hemodynamic factor influencing erythrocyte aggregation. An increase in shear rate can break up the formation of rouleaux and reduce the intensity of the echogenicity of flowing blood. In contrast to steady flow, some studies have shown the existence of a cyclic variation in the echogenicity of pulsatile flowing blood both in vitro and in vivo (Cloutier and Shung 1993; De Kroon et al. 1991; Missaridis and Shung 1999; Paeng and Shung 2003; Paeng et al. 2001, 2010; Huang 2009, 2011; Thompson et al. 1985; Wu and Shung 1996). In general, the cyclic variations in the ultrasonic backscatter signal or Doppler power have been clearly observed for blood with a higher hematocrit, lower stroke rate, and lower peak flow velocity, and when using highfrequency ultrasound. It was proposed that the shear rate was responsible for the cyclic variation in the echogenicity of blood under pulsatile flow conditions in a tube when measured with spatial and temporal variations in ultrasonic Doppler signals. Because the shear rate acting on the erythrocyte aggregates across the tube varies with the time during a flow cycle, cyclic variations in echogenicity might be related to the dynamics of erythrocyte aggregation (Lin and Shung 1999). However, recent results indicate that the echogenicity of flowing blood is minimized in the late diastole phase and increases during the early systole phase, reaching a maximum before peak systole. It subsequently decreases again, reaching a minimum at late diastole (Huang 2009, 2011; Paeng et al. 2001). This observation opposes the previously held opinion that the increase in shear rate should disrupt erythrocyte aggregation and reduce the echogenicity of blood; however, it was found that the echogenicity increased during the acceleration phase in pulsatile flow. Consequently, some studies indicate that the higher echogenicity of blood during early systole cannot be explained by shear-rate analysis alone, which led to the hypothesis that flow acceleration underlies the increase in echogenicity of blood during the acceleration phase. Flow acceleration might promote rouleaux formation, because when the flow is accelerating there will be more opportunities for erythrocytes to interact to form larger rouleaux (Cao et al. 2001; Paeng et al. 2001, 2004). Although the cyclic variation in echogenicity was explained by combined effects on erythrocyte aggregates, the contribution of each factor to the echogenicity of pulsatile flowing blood remains unclear because it is difficult to estimate the effect of these two hemodynamic factors independently. In other words, the interpretation of the effects of these mechanisms on erythrocyte aggregates needs further detailed inspection to obtain a better understanding of their influence. In addition, in most previous studies the Doppler power or backscatter power was measured to

671

represent the level of erythrocyte aggregation; however, the effect of attenuation compensation on backscatter measurements was not considered. Because the attenuation of ultrasound in blood depends on the shear rate (Huang and Chang 2011), the effect of attenuation on the cyclic variation of echogenicity should also be considered under conditions of pulsatile flow. The purpose of this study was to determine quantitatively how flow acceleration affects the cyclic variation of blood echogenicity under pulsatile flow conditions. To simplify the complicated kinetics of blood flowing in a circular tube, a Couette flow apparatus was designed to simulate the pulsatile flow condition (Nguyen et al. 2008). Fresh porcine whole blood and erythrocyte suspensions with hematocrit values of 20% and 40%, respectively, were circulated in the flow phantom at different flow acceleration settings. High-frequency ultrasound is more sensitive for detecting erythrocyte aggregates (Huang 2009); therefore, the ultrasonic backscattering coefficient and flow velocity of flowing blood were measured synchronously using a 35-MHz ultrasonic transducer and a 35-MHz pulsed-wave Doppler flowmeter, respectively. To measure the backscattering coefficients accurately, backscatter data were compensated according to the shear-rate–dependent attenuation. This study focused on how blood flow acceleration affects erythrocyte aggregates and the echogenicity during the acceleration phase under pulsatile flow conditions. MATERIALS AND METHODS Blood samples All experiments were performed on porcine blood within 24 h of its collection (from a local slaughterhouse). Whole blood (1 L, to which 30 mL of the anticoagulant ethylenediaminetetraacetic acid was added at a concentration of 11 gm/dL) was passed through a sponge to filter out impurities such as fatty tissue and hair. It was then centrifuged and washed twice using a saline buffer solution to separate the erythrocytes from the plasma and other cells. The concentrated erythrocytes and plasma were stored in a refrigerator, and the desired hematocrit values for the whole-blood experiments (20% and 40%) were obtained by mixing the appropriate amount of concentrated erythrocytes with plasma. Erythrocyte suspensions with hematocrit values of 6% and 40% in 0.9% saline solution were also prepared. The erythrocyte suspension with a 6% hematocrit was used as a reference medium for measuring the backscattering coefficient. In total, 10 porcine blood samples from 10 different animals were used in the present study. Couette flow system A Couette flow system was designed for this study to simulate the blood flowing under pulsatile flow

672

Ultrasound in Medicine and Biology

Volume 39, Number 4, 2013

Fig. 1. Block diagram of the experimental system and illustration of the Couette flow system. ADC 5 high resolution digitizer; PC 5 personal computer.

conditions. The blood sample was placed in the 2-mm gap between two parallel disks that were moving relative to each other. The Couette flow system and its mechanism of action are described in detail elsewhere (Huang and Chang 2011). Two symmetrical circular acoustic windows with a diameter of 13 mm were made in the upper disk such that the distance between the center of the disk and each acoustic window was 40 mm. These two acoustic windows were covered with a polymethylpentene film (TPX; Mitsui, White Plains, NY, USA), which allowed the passage of most of the ultrasound energy transmitted to and received from the blood. A 35-MHz focused ultrasonic transducer (National Institutes of Health Resource on Medical Ultrasonic Transducer Technology, University of Southern California, Los Angeles, CA, USA) was positioned perpendicular to one acoustic window for attenuation and backscattering coefficient measurements. Another 35-MHz transducer was placed with a Doppler angle of 60 degrees to the other acoustic window for flow velocity measurements. Both transducers were immersed in distilled water. For all measurements, the sample volume of each transducer was located near the focal zone. The rotation speed was controlled by a personal computer via the printer-port interface to the motor controller. Two types of acceleration pattern were set to simulate pulsatile flow in this study: 1. The flow velocity was increased gradually from stationary to a constant velocity, and then different instantaneous accelerations were applied to the blood to create peak flow velocities of 0.3, 0.6, and 1.2 m/s. Under these conditions, the blood flow was stopped

for 1 s after the continuous motion of flow for 1, 2 and 3 s (designated as A1S1, A2S1 and A3S1, respectively). Before applying the instantaneous acceleration, the initial velocity change was the same within 0.5, 1 and 2 s, corresponding to continuous motion of flow for 1, 2 and 3 s, respectively. 2. The blood flow was set at a constant initial velocity of 0.07, 0.14 and 0.26 m/s for 3 s, and then the instantaneous accelerations were applied to blood within 1 s, providing peak flow velocities of 0.4, 0.5 and 0.6 m/s, respectively. The velocity was subsequently returned to the former velocity for a further 3 s. Experimental setup for measuring backscatter Figure 1 shows the experimental setup for measuring backscatter. The 35-MHz focused transducers were excited by ten cycles of high-voltage sinusoidal tone bursts produced by a power amplifier (25A250, Amplifier Research, Souderton, PA, USA). The 35-MHz tone-burst signals were generated at a pulse repetition frequency (PRF) of 100 Hz by a high-frequency function generator (AFG3102, Tektronix, Beaverton, OR, USA), which was controlled by a personal computer via a USB interface. A protection circuit comprising an expander (DEX-3; Matec, Hopkinton, MA, USA) and a limiter (DL-1, Matec) was placed in front of the transducer. The backscatter signals were received and amplified with a 30-dB preamplifier (LN1000, Amplifier Research) and then filtered by a 35-MHz bandpass filter (Mini-Circuits, Brooklyn, NY, USA). The PRF trigger of the function generator was used to synchronize the collection of the echoes at a sampling frequency of 200 MHz by a high-resolution

Flow acceleration on blood aggregation d C.-C. HUANG et al.

(12-bit) digitizer (PCI-5124; National Instruments, Austin, TX, USA). The spatial peak temporal average intensity and electrical power of measurement system were 5.2 mW/cm2 and 200 mW, respectively. The reflection echoes from the lower disk were acquired in a 1.0-ms time window for calculating the attenuation coefficient of flowing blood. After recording the stronger echoes, the voltage range of the digitizer was changed to acquire the backscattered signals from the blood in a 2.5-ms time window corresponding to a volume of interest located within the gap between the two parallel disks. The attenuation coefficient (a) was measured by the substitution method (Huang and Chang 2011) according to: a5

20 log10 ðA1 ð fÞ=A2 ð fÞÞ 2d

(1)

where d is the distance between the two parallel disks in the Couette flow system, and A1(f) and A2(f) represent the amplitudes of the spectra of the echoes received from the lower disk in distilled water and from the disk containing a blood sample, respectively. The attenuation coefficients were measured over flow cycles. The modified substitution approach was used for backscattering coefficient measurements (Huang 2010). The backscattering coefficient (h) of a blood sample can be expressed as:   P 24Lða0 2aÞ h 5 h0 e (2) P0

673

Table 1. The specifications of Doppler flow meter Specification Setting Center operation frequency Pulse repetition frequency Transmitted pulse duration Wall filter

35 MHz 140 kHz 0.285 ms (10 cycles) 50 Hz

that the timing was controlled by an FPGA module (EP2C5T144; Altera, San Jose, CA, USA) and some high-speed electronic components—such as filters and a demodulator (MIQC-60D, Mini-Circuits)—were used for the 35-MHz flowmeter in this study. The specifications of this Doppler flowmeter are listed in Table 1. The sample volume was located at the center of the gap. The directional Doppler signal was recorded by the same digitizer at a sampling frequency of 500 kHz using another channel. The Doppler signal was segmented using a temporal window of 5000 sample points. The Doppler spectrum was obtained using the fast Fourier transform (FFT), and the overlap between consecutive processes was 50%. A 1024-point FFT was applied to the Hamming-windowed signal from each segment. The Doppler spectrogram corresponding to each temporal window was then displayed in grey level. The flow velocity (v) within the sample volume was calculated as: v5

fd c 2 f0 cos q

(3)

where the h0 and a0 respectively denote the backscatter and attenuation coefficient of the 6%-hematocrit erythrocyte suspension, a is the shear-rate-dependent attenuation of the blood sample, P0 and P respectively represent the backscatter power from the 6%-hematocrit erythrocyte suspension and the blood sample in the same insonified region under the same measurement condition, and L is the distance between the acoustic window and the center of the sample volume. The backscattering and attenuation coefficients for 6%-hematocrit porcine erythrocyte suspensions at a frequency of 35 MHz were referred from a previous study (Wang and Shung 1997). The measured attenuation n and backscattering coefficients were low-pass filtered using a moving-average algorithm to temporally smooth the data.

where fd is the Doppler shift frequency, f0 is the center frequency of the transmitted ultrasound, c is the speed of sound in blood (1550 m/s) and q is the angle between the acoustic beam and the blood flow. A total of 50 pulsatile cycles was recorded and averaged to represent the ensemble cyclic variation of attenuation, echogenicity, and velocity waveform as functions of time. The stepper motor of the Couette flow system, function generator and digitizer were commanded synchronously by a personal computer using LabView software (National Instruments). Data analyses were performed on a personal computer using MATLAB (Math Works, Natick, MA, USA). All measurements were performed at a room temperature of 25.0 6 0.5 C.

Experimental setup for measuring the Doppler velocity A high-frequency (35 MHz) pulsed-wave Doppler flowmeter was developed for flow velocity measurements, including an FPGA module, bipolar pulser, RF amplifier, bandpass filters, demodulator, sample and hold circuit, PRF filter, wall filter, and audio amplifier. The Doppler flowmeter used in our 10-MHz system is described in detail elsewhere (Huang et al. 2012), except

Figure 2a shows typical Doppler spectrograms for 40%-hematocrit whole blood with the A2S1 acceleration pattern in the Couette flow system. Peak flow velocities of 0.3, 0.6, and 1.2 m/s were reached in this setting. The Doppler spectra exhibited a clear window during the acceleration phase in all experiments. In other words, no turbulent flow was generated in the Couette flow system. Figure 2 (b–d) shows the average results of

RESULTS

674

Ultrasound in Medicine and Biology

Volume 39, Number 4, 2013

Fig. 2. (a) Typical Doppler spectrograms for 40% hematocrit whole blood at an acceleration pattern of A2S1 for three different peak flow velocities of 0.3, 0.6, and 1.2 m/s (from left to right). The average results of Doppler velocity over a flow cycle under varied acceleration patterns with different peak flow velocities in the first acceleration setting: (b) A1S1, (c) A2S1 and (d) A3S1.

Doppler velocity over the flow cycle under different acceleration patterns (40% hematocrit whole blood). In each setting, the initial changes of velocity were the same within a constant duration, and then the instantaneous accelerations were applied to blood. Figure 3 shows the attenuation coefficient, backscattering coefficient and flow velocity results over an A2S1 cycle for 40% hematocrit whole blood. The mean and standard deviation values of 10 different blood samples are shown as lines and error bars, respectively. The cyclic variation in backscattering coefficient is clearly observed in the figure, whereas no such variation is evident in the attenuation. Analysis of the effect of acceleration on erythrocyte aggregation required the definition of some parameters. Referring to Figure 3, D is the difference between the minimum and the maximum backscattering coefficients in a cycle, t and V are the time and its corresponding flow velocity of the peak backscattering coefficient, respectively, and average is the mean value of all backscattering coefficients over a cycle. The values of attenuation were 1.1 6 0.12 and 1.5 6 0.18 dB/mm for the 20% and 40% hematocrit whole blood in call cases, respectively. These values are consistent with those measured in our previous study (Huang and Chang 2011); however, the attenuation curve for each flow cycle was not affected by the acceleration pattern. The measured attenuation coefficients were used to compute the backscattering coefficients in equation 2.

At each data point, the backscattering coefficient was compensated using the measured attenuation coefficient at the corresponding flow velocity. Figures 4 and 5 show the backscattering coefficients and velocity waveforms over a cycle under different acceleration patterns for 20% and 40% hematocrit whole blood, respectively. The standard deviations of the backscatter curves are not plotted in the figure in order to increase the clarity of the curves. The cyclic variations of the backscatter were observed clearly at 40% hematocrit, and the peaks of the backscattering coefficient curve were found to occur before the peaks of the velocity waveform in all

Fig. 3. The measured results of attenuation coefficient, backscattering coefficient, and flow velocity over a A2S1 flow cycle for 40% hematocrit whole blood. The parameters of D, t and V are marked in the figure.

Flow acceleration on blood aggregation d C.-C. HUANG et al.

675

Fig. 4. The mean values of backscattering coefficients and velocity waveforms for 20% hematocrit whole blood over a flow cycle at three different acceleration patterns (A1S1, A2S1 and A3S1, from left to right) for three different peak flow velocities (0.3, 0.6, and 1.2 m/s from top to bottom).

cases. The backscattering coefficient increased during the early flow cycle, subsequently gradually reducing before the instantaneous acceleration phase. The values of the average, D, t, and V parameters for 20% and 40% hematocrit whole blood over a cycle are listed in Tables 2 and 3, respectively. The mean and standard deviation values are also listed in the tables. Figure 6 shows the backscattering coefficient curves and velocity waveforms over a cycle under different acceleration patterns for the 40% hematocrit erythrocyte suspension.

No cyclic variations of backscatter signals were apparent for the erythrocyte suspension. Figure 7 shows typical Doppler spectrograms of 40% hematocrit whole blood for the second experimental setting, in which an instantaneous acceleration was applied to the flowing blood. The spectra were wider at higher flow velocities. The average Doppler velocities as functions of the flow cycle under different acceleration patterns are also shown in the figure. The attenuation values were 1.5 6 0.17 and 1.0 6 0.08 dB/mm for

Fig. 5. Same as Figure 4, except that the data were obtained for 40% hematocrit whole blood.

676

Ultrasound in Medicine and Biology

Volume 39, Number 4, 2013

Table 2. Summary of measured parameters of cyclic variation for 20% hematocrit whole blood under different acceleration patterns Acceleration 1 s, stop 1 s (A1S1) Peak velocity (m/s) 0.3 0.6 1.2

average*

D

t (s) V (m/s)

4363 6 1704 — 3911 6 1601 654 6 282 3324 6 1248 692 6 351

— 0.6 0.7

— 0.07 0.08

Acceleration 2 s, stop 1 s (A2S1) average

D

4335 6 1613 — 4014 6 1476 649 6 311 3598 6 1594 709 6 382

t (s) V (m/s) — 1.2 1.3

— 0.08 0.09

Acceleration 3 s, stop 1 s (A3S1) average

D

3819 6 1104 — 3466 6 1100 — 3073 6 1108 651 6 222

t (s) V (m/s) — — 2.1

— — 0.12

— 5 no observable difference. * The unit for average and D is 106 cm21sr21.

the 40% hematocrit whole blood and erythrocyte suspension for the second experimental setting, respectively. Again the attenuation for each flow cycle was not affected by the acceleration pattern. Figure 8 shows the backscattering coefficients and velocity waveforms over a single cycle under different acceleration patterns for the 40% hematocrit whole blood and erythrocyte suspension, respectively. The standard deviations of the backscatter curves are not plotted in the figure in order to make the curves clearer. In whole blood, the backscattering coefficients seemed to remain constant during the first 3 s and then reduce rapidly after instantaneous acceleration was induced in the blood. Backscattering coefficients subsequently increased gradually as the flow velocity returned to the initial level. No variations of backscatter were apparent for the erythrocyte suspension. DISCUSSION It is well known that ultrasonic backscatter from scattering elements is influenced by several physical parameters, such as the size, shape, density, compressibility, concentration and distribution of the scatterers. The formation of larger rouleaux in whole blood will increase the backscatter; therefore, measurement of the backscatter signal from whole blood can be used to detect the level of erythrocyte aggregation. In our previous study (Huang and Chang 2011), we found that the attenuation of ultrasound in blood depends on the shear rate. Because attenuation compensation is an important factor for backscatter measurements, we doubted whether the cyclic variation of echogenicity was caused by the variation of

ultrasonic attenuation. However, this conjecture was not supported by the experimental results, in which the attenuation coefficient for each flow cycle remained constant irrespective of the acceleration pattern. The measurements of attenuation coefficients were taken from flowing blood, for which the flow velocity was maintained for 50 s (Huang and Chang 2011). However, the instantaneous change of flow velocity was achieved by 1 s in present study. In other words, instantaneous variations in flow velocity did not affect the ultrasonic attenuation in pulsatile flowing blood. This result contradicts our previous finding of a decrease in the attenuation coefficient of blood with increasing shear rate. This discrepancy may be explained by the time taken for blood to change its viscosity. Because the ultrasonic attenuation varies little during a pulsatile cycle, it is safe to assume that the cyclic variation of the backscatter signal from pulsatile flow was associated mainly with the properties of the scattering elements. Cyclic variations in echogenicity were clearer in 40% hematocrit than in 20% hematocrit whole blood, as shown in Figures 4 and 5. In some cases, it was difficult to identify the cyclic variation for 20% hematocrit whole blood, particularly at lower peak-flow velocities. This finding is in a good agreement with previous studies involving flowing blood in a tube finding that cyclic variations of echogenicity were hematocrit dependent (Huang 2009, 2011). In addition, the average values decreased with increasing peak flow velocity for both hematocrit values (Tables 2 and 3). This observation is also consistent with those of previous studies (Huang 2009, 2011), although our experiments were performed

Table 3. Summary of measured parameters of cyclic variation for 40% hematocrit whole blood under different acceleration patterns Acceleration 1 s, stop 1 s (A1S1) Peak velocity (m/s) 0.3 0.6 1.2

average*

D

4673 6 1904 688 6 238 4042 6 1481 881 6 321 3473 6 1504 979 6 332

t (s) V (m/s) 0.4 0.5 0.6

* The unit for average and D is 106 cm21 sr21.

0.05 0.06 0.07

Acceleration 2 s, stop 1 s (A2S1) average

D

4507 6 1288 997 6 320 4005 6 1309 1033 6 385 3524 6 1287 1239 6 420

t (s) V (m/s) 0.8 0.9 1.0

0.05 0.06 0.07

Acceleration 3 s, stop 1 s (A3S1) average

D

4822 6 1592 1602 6 495 4396 6 1690 1616 6 421 3767 6 1560 1710 6 462

t (s) V (m/s) 1.0 1.0 1.2

0.05 0.05 0.07

Flow acceleration on blood aggregation d C.-C. HUANG et al.

677

Fig. 6. Same as Figure 4, except that the data were obtained for 40% erythrocyte suspension.

using a Couette flow system. Our finding that the peak backscattering coefficient curve occurred before the peak of the velocity waveform, increasing during the early motion of flow, is also similar to previous results (Huang 2009, 2011; Lin and Shung 1999; Paeng et al. 2001; Wu and Shung 1996), as shown in Figures 4 and 5. This result appears to be consistent with a previous hypothesis attributing the cyclic variations of echogenicity to the combined effects of shear rate and acceleration. However, a contradictive phenomenon occurred as

instantaneous acceleration was applied to the flowing blood; the echogenicity was not enhanced during the instantaneous acceleration phase, and was instead reduced to near the minimum in all cases. According to the previous hypothesis (Cao et al. 2001; Paeng et al. 2001, 2004), acceleration may enhance the probability of erythrocyte collision to form larger rouleaux during the acceleration phase. However, the results of present study seem to imply that flow acceleration does not promote the formation of rouleaux, even though the blood

Fig. 7. Typical Doppler spectrograms and the corresponding Doppler velocity waveforms over a flow cycle for second acceleration setting. The initial flow velocities are 0.07, 0.14 and 0.26 m/s from top to bottom. Blood sample is 40% hematocrit whole blood.

678

Ultrasound in Medicine and Biology

reached a higher peak velocity. Regardless of the duration of the motion alterations, the rouleaux were broken up by the instantaneous acceleration of flow, as revealed by the decrease in echogenicity as the flow reached a certain velocity. Comparison of different acceleration patterns revealed that the value of t increased with the peak velocity and motion duration of the flowing blood. For example, the values of t for 40% hematocrit whole blood at a peak velocity of 0.6 m/s were 0.5, 0.9, and 1.0 s for A1S1, A2S1 and A3S1, respectively (Table 3), whereas the V value was constant (at 0.06 m/s) for these three patterns. Similar situations were observed for other peak velocities. In other words, regardless of how much velocity was accelerated, the erythrocyte aggregation was disrupted as the flow velocity reached 0.05–0.07 m/s. These velocities are close to those measured in our previous study for 40% hematocrit blood, in which a threshold velocity of 0.08 m/s was required for erythrocyte disaggregation under conditions of pulsatile flow (Huang 2011). Flow turbulence can be considered as another factor that enhances echogenicity, but this occurred only in an erythrocyte suspension (Shung et al. 1992). Figure 6 shows the experimental results for the 40% hematocrit erythrocyte suspension under different acceleration patterns; no apparent cyclic variations of echogenicity were observed in any of the cases. It is therefore safe to assume that the flow turbulence exerts only minor effects on cyclic variations of echogenicity. This result represents evidence that the cyclic variations of echogenicity can be attributed to erythrocyte aggregation because it occurs only in whole blood. To further confirm the effect of acceleration on erythrocyte aggregation, an instantaneous acceleration was applied to the flowing blood at three different initial velocities (Fig. 7). The measured attenuation coefficients remained constant for both whole blood and the erythrocyte suspension over the flow cycle, which eliminates any effect of attenuation on cyclic variations of echogenicity. However, the patterns of the backscattering coefficient curve differed from those in previous experiments. The echogenicity decreased rapidly with the application of instantaneous acceleration to 40% hematocrit flowing whole blood (Fig. 8), which represents further evidence that the instantaneous acceleration did not promote erythrocyte aggregation. Instead, the larger rouleaux in flowing blood were disrupted to form smaller rouleaux because of the acceleration. Because the rouleaux are initially larger under lower flow velocities (Huang and Wang 2007), the decrease in echogenicity was stronger for a lower initial velocity. Again, no cyclic variations of echogenicity were apparent in erythrocyte suspensions, because erythrocytes do not aggregate in saline solution.

Volume 39, Number 4, 2013

In the results compared with the previous acceleration setting, there was no apparent increase in echogenicity from whole blood before the acceleration of flow (Fig. 8). This discrepancy can be attributed to the baseline of flowing blood being under a stationary or moving condition. For most in vitro and in vivo studies, pulsatile flowing blood initially has a velocity close to zero, and then accelerates and decelerates due to systole and diastole, respectively. The increase in echogenicity that occurs during the early systole phase observed in most other studies was also observed under our first acceleration setting, although only a slow flow was applied to the initially stationary whole blood. If the acceleration is not a major contributor to cyclic variations in echogenicity, we believe the main factors to be the effect of shear rate and the distribution of erythrocytes on aggregation. According to previous studies of steady flow (Cloutier et al. 1996; Shehada et al. 1994), erythrocyte aggregation is promoted predominantly at a shear rate of 0.1–10 s21 in porcine blood. This finding might explain the increase in echogenicity during pulsatile flow, because aggregation is enhanced when stationary blood starts moving. However, the echogenicity did not increase under the second setting because the initial shear rate would have been higher than this range. As the motion of erythrocytes continued to accelerate, the alignment and rotation of the rouleaux can change rapidly with flow because of the nonuniform shear force applied to the blood. This change in the spatial distribution of erythrocytes also increases the echogenicity during pulsatile flow (Huang 2011). In addition, the values of D increased at the peak flow velocity (Table 3), which means that the larger rouleaux were strongly disrupted by the higher shear rates. Therefore, there is no question that the echogenicity decreased after the commencement of blood flow up to a certain velocity. The contemporary literature indicates that backscattering from blood is dependent on the size, shape, compressibility, and density of the scatterers for a given ultrasound frequency (Shung and Thieme 1993). If the ultrasonic measurements were performed in the same blood sample, the backscattering signal from pulsatile flow is dependent on the ultrasound frequency (Huang 2009; Huang and Chang 2011). In other words, the experimental results in the present study might be influenced by using different ultrasound frequencies. However, the sensitivity and resolution of 35-MHz ultrasound are sufficient for detecting the red blood cell aggregation under the pulsatile flow (Huang 2009; Nguyen et al. 2008). In addition, the best way to understand the effect of acceleration on the cyclic variations of blood echogenicity is using animal experiments and clinical trials. However, it is difficult to control the blood flow conditions such as peak flow velocity, acceleration, vessel wall compliance, hematocrit and heart rate in a realistic pulsatile

Flow acceleration on blood aggregation d C.-C. HUANG et al.

679

Fig. 8. The mean values of backscattering coefficients and velocity waveforms for 40% hematocrit whole blood and erythrocyte suspension over a flow cycle at three different initial flow velocities (0.07, 0.14 and 0.26 m/s, from top to bottom).

flow in vivo. On the contrary, these variables are easily controlled in Couette flow for studying the specific factors that influence the echogenicity from flowing blood. Currently, we cannot say that our findings are in full agreement with clinical environment. In the future works, an ex vivo study that takes a vessel out of animal and pump the blood into circulation will be performed to verify the relationship between blood echogenicity and red blood cell aggregation at different flow conditions. CONCLUSIONS The cyclic variation of echogenicity under pulsatile flow has been validated to relate with the erythrocyte aggregation both in vitro and in vivo. However, the mechanism underlying the increase in echogenicity during the early systolic phase remains unclear. Because the temporal variation in echogenicity over a flow cycle cannot be explained completely by shear rate alone, the effect of flow acceleration was hypothesized as another factor that enhances aggregation. To confirm this hypothesis, different instantaneous acceleration patterns were applied to flowing blood using a Couette flow system in this study. The effect of ultrasonic attenuation on the cyclic variation in echogenicity was eliminated because there were only small variations in the attenuation coeffi-

cient over a flow cycle. Cyclic variations in echogenicity were clearly observed in the experiments with whole blood at a higher hematocrit of 40%, and no variations in echogenicity were apparent for the erythrocyte suspension. However, the echogenicity was not enhanced with the application of an instantaneous acceleration to flowing blood in all cases. This experimental result represents important evidence that the instantaneous acceleration does not promote erythrocyte aggregation in whole blood, even with the application of a higher peak velocity. Compared with stationary- and moving-blood experiments, the cyclic variation of echogenicity can be attributed mainly to the effect of shear rate and the distribution of erythrocytes on aggregation. In porcine whole blood with a hematocrit of 40%, a threshold range of velocity 0.05–0.07 m/s is needed for erythrocyte disaggregation under pulsatile flow conditions. Acknowledgments—This work was supported by the National Science Council of Taiwan under grant NSC 99-2221-E-030-001-MY2. The authors thank Mr. Y. C. Chang for assistance with the experiments.

REFERENCES Bosch T, Wendler T, Jaeger BR, Samtleben W. Improvement of hemorheology by DALI apheresis: Acute effects on plasma viscosity and erythrocyte aggregation in hypercholesterolemic patients. Ther Apher 2001;5:372–376.

680

Ultrasound in Medicine and Biology

Cao P-J, Paeng D-G, Shung KK. The ‘‘black hole’’ phenomenon in ultrasonic backscattering measurement under pulsatile flow with porcine whole blood in a rigid tube. Biorheology 2001;38:15–26. Chabanel A, Horellou MH, Conard J, Samama MM. Red blood cell aggregability in patients with a history of leg vein thrombosis: influence of post-thrombotic treatment. Br J Haematol 1994;88:174–179. Cloutier G, Shung KK. Study of red cell aggregation in pulsatile flow from ultrasonic Doppler power measurements. Biorheology 1993; 30:443–461. Cloutier G, Qin Z, Durand LG, Teh BG. Power Doppler ultrasound evaluation of the shear rate and shear stress dependence of red blood cell aggregation. IEEE Trans Biomed Eng 1996;43:441–450. Cloutier G, Weng X, Roederer GO, Allard L, Tardif F, Beaulieu R. Differences in the erythrocyte aggregation level between veins and arteries of normolipidemic and hyperlipidemic individuals. Ultrasound Med Biol 1997;23:1383–1393. Cloutier G, Zimmer A, Yu FTH, Chiasson JL. Increased shear rate resistance and fastest kinetics of erythrocyte aggregation in diabetes measured with ultrasound. Diabetes Care 2008;31:1400–1402. Demiroglu H, Barista I, Dundar S. Erythrocyte aggregability in patients with coronary heart disease. Clin Hemorheol 1996;16:313–317. De Kroon MGM, Slager CJ, Gussenhoven WJ, Serruys PW, Roelandt JR, Bom N. Cyclic changes of blood echogenicity in high frequency ultrasound. Ultrasound Med Biol 1991;17:723–728. Hahn R, Muller-Seydlits PM, Jockel KH, Hubert H, Heimburg P. Viscoelasticity and red blood cell aggregation in patients with coronary heart disease. Angiology 1989;40:914–920. Huang CC, Wang SH. Statistical variations of ultrasound signals backscattered from flowing blood. Ultrasound Med Biol 2007;33: 1943–1954. Huang CC, Lin YH, Wang SH. The effect of kinetic properties on statistical variations of ultrasound signals backscattered from flowing blood. Jan J Appl Phys 2009;48:027002–1-02002-7. Huang CC. Cyclic variations of high-frequency ultrasonic backscattering from blood under pulsatile flow. IEEE Trans Ultrason Ferroelec Freq Contr 2009;56:1677–1688. Huang CC. High frequency attenuation and backscatter measurements of rat blood between 30-60 MHz. Phys Med Biol 2010;55: 5801–5816. Huang CC, Chang YC. Ultrasonic attenuation and backscatter from flowing whole blood are dependent on shear rate and hematocrit between 10 and 50 MHz. IEEE Trans Ultrason Ferroelec Freq Contr 2011;58:357–368. Huang CC. Detecting spatial variations of erythrocytes by ultrasound backscattering statistical parameters under pulsatile flow. IEEE Trans Biomed Eng 2011;58:1163–1171. Huang CC, Lee PY, Chen PY, Liu TY. Design and implementation of a smart-phone based portable ultrasound pulse-wave Doppler device for blood flow measurement. IEEE Trans Ultrason Ferroelec Freq Contr 2012;59:182–188. Le Devehat C, Vimeux M, Bondoux G, Khodabandehlou T. Red blood cell aggregation in diabetes mellitus. Int Angiol 1990;9:11–15. Lin YH, Shung KK. Ultrasonic backscattering from porcine whole blood of varying hematocrit and shear rate under pulsatile flow. Ultrasound Med Biol 1999;25:1151–1158.

Volume 39, Number 4, 2013 Li Y, Bok TH, Yang JH, Choi MJ, Paeng DG. The acute effects of smoking on the cyclic variations in blood echogenicity of carotid artery. Ultrasound Med Biol 2011;37:513–521. Missaridis TX, Shung KK. The effect of hemodynamics, vessel wall compliance and hematocrit on ultrasonic Doppler power: An in vitro study. Ultrasound Med Biol 1999;25:549–559. Nguyen LC, Yu FTH, Cloutier G. Cyclic changes in blood echogenicity under pulsatile flow are frequency dependent. Ultrasound Med Biol 2008;34:664–673. Paeng DG, Cao PJ, Shung KK. Doppler power variation from porcine blood under steady and pulsatile flow. Ultrasound Med Biol 2001; 27:1245–1254. Paeng DG, Shung KK. Cyclic and radial variation of the Doppler power from porcine whole blood. IEEE Trans Ultrason Ferroelec Freq Contr 2003;50:614–622. Paeng DG, Chiao RY, Shung KK. Echogenicity variations from porcine blood I: The ‘‘bright collapsing ring’’ under pulsatile flow. Ultrasound Med Biol 2004;30:45–55. Paeng DG, Nam KH, Shung KK. Cyclic and radial variation of echogenicity of blood un human carotid arteries observed by harmonic imaging. Ultrasound Med Biol 2010;36:1118–1124. Razavian SM, Atger V, Giral P, Cambillau M, Del-Pino M, Simon AC, Moatti N, Levenson J. Influence of HDL subfractions on erythrocyte aggregation in hypercholesterolemic men. PCVMETRA group. Arterioscler Thromb 1994;14:361–366. Razavian SM, Del Pino M, Simon A, Levenson J. Increase in erythrocyte disaggregation shear stress in hypertension. Hypertension 1992;20:247–252. Samocha-Bonet D, Ben-Ami R, Shapira I, Shenkerman G, Abu-Abeid S, Stern N, Mardi T, Tulchinski T, Deutsch V, Yedgar S, Barshtein G, Berliner S. Flow-resistant red blood cell aggregation in morbid obesity. Int J Obesity 2004;28:1528–1534. Schmid-Sch€onbein H, Volger E. Red-cell aggregation and red-cell deformability in diabetes. Diabetes 1976;25:897–902. Shehada REN, Cobbold RSC, Mo LYL. Aggregation effects in whole blood: Influence of time and shear rate measured using ultrasound. Biorheology 1994;1:115–135. Shung KK, Cloutier G, Lim CC. The effects of hematocrit, shear rate and turbulence on ultrasonic Doppler spectrum from blood. IEEE Trans Biomed Eng 1992;39:462–469. Shung KK, Thieme GA. Ultrasonic scattering in biological tissues. Boca Raton, FL: CRC Press; 1993. Thompson SR, Trudinger BJ, Cook CM. Doppler ultrasound waveforms in the fetal umbilical artery: quantitative analysis technique. Ultrasound Med Biol 1985;11:707–718. Wang SH, Shung KK. An approach for measuring ultrasonic backscattering from biological tissues with focused transducer. IEEE Trans Biomed Eng 1997;44:549–554. Wu SJ, Shung KK. Cyclic variation of Doppler power from whole blood under pulsatile flow. Ultrasound Med Biol 1996;22:883–894. Yuan YW, Shung KK. Ultrasonic backscatter from flowing whold blood I: Dependence on shear rate and hematocrit. J Acoust Soc Am 1988a;84:52–58. Yuan YW, Shung KK. Ultrasonic backscatter from flowing whold blood II: Dependence on frequency and fibrinogen concentration. J Acoust Soc Am 1988b;84:1195–1200.