MEMS Thermal Sensors to Detect Changes in Heat Transfer in the Pre-Atherosclerotic Regions of Fat-Fed New Zealand White Rabbits
- 253 Downloads
Real-time detection of pre-atherosclerotic regions remains an unmet clinical challenge. We previously demonstrated the application of micro-electro-mechanical systems (MEMS) to detect changes in convective heat transfer in terms of sensor output voltages in the zone of flow reversal in an in vitro stenotic model. We hereby demonstrated changes in sensor output voltages in the pre-atherosclerotic regions in the New Zealand White rabbits fed on hypercholesterolemic diet (HD). After 8 weeks, we observed that mean output voltages (Vave) were similar in the distal aortic arch, thoracic, and abdominal aortas in the normal standard diet (ND) group, consistent with an absence of atherosclerosis. In HD group, Vave increased in the distal aortic arch (HD: Vave = 1.05 ± 0.04 V; ND: Vave = 0.12 ± 0.01 V, n = 3, p < 0.05) and in the thoracic aortas (HD: Vave = 0.72 ± 0.06 V; ND: Vave = 0.13 ± 0.024 V, n = 3, p < 0.05), consistent with the histological presence of pre-atherosclerosis. Despite HD diet, Vave magnitudes were similar to ND group in the abdominal aortas (HD: Vave = 0.14 ± 0.003 V; ND: Vave = 0.14 ± 0.004 V, n = 3), corroborating histological absence of pre-atherosclerosis. Hence, MEMS thermal sensors provide a new approach to detect changes in convective heat transfer in the pre-atherosclerotic regions.
KeywordsMEMS Atherosclerosis Heat transfer Disturbed flow Shear stress
Mounting evidence supports that hemodynamics, namely, fluid shear stress, is intimately involved in vascular oxidative stress,24 inflammatory responses, and atherosclerosis.7,10,26,34 In the athero-prone regions, flow separation, and secondary flow develop. In response to cardiac contraction, the migrating stagnation points generate low and oscillating flow, otherwise known as disturbed or secondary flow.11 However, real-time detection of changing flow profiles in the presence of pre-atherosclerotic regions remains a clinical challenge.
Intravascular strategy has emerged to address mechanically unstable plaque. In the arterial regions where flow separation develops, low and oscillatory wall shear stress predict the progression of atherosclerosis,30 vessel wall remodeling,9 and development of abdominal aortic aneurysm.27,33 The combination of three dimensional intravascular ultrasound (IVUS) images and finite element analysis have been reported to assess plaque size, shape, expansive remodeling, calcification, and lipid core.18 Stone et al. demonstrated integration of IVUS, biplane coronary angiography, and coronary blood flow measurements to elucidate the effects of shear stress on the progression of atherosclerosis, vascular remodeling, and in-stent re-stenosis in humans.21,30 Jaffer et al. further demonstrated intravascular deployment of near-infrared fluorescence catheter to detect proteolytically active atherosclerotic lesions in New Zealand White (NZW) rabbits.19 In our group, we introduced the micro-electro-mechanical systems (MEMS) thermal sensors to provide high spatial (1/10th of a string of hair) and temporal resolution (kHz)2,3,36 that enables us to link shear stress with pre-atherosclerotic lesions.
The operational principle of the MEMS sensors is based on convective cooling of a heated sensing element as fluid flows over the sensing element surface. The heat transfer from the heated surface to the fluid depends on the flow characteristics in the viscous region of the boundary layer.12 In response to instantaneous fluctuation in temperature, the heat dissipation from the sensors to the blood flow alters the resistance of the sensing elements, from which voltage changes are acquired.15 Using a stenotic model, we reported that the MEMS sensors detected changes in convective heat transfer in the throat of stenosis and regions of flow reversal.4
To demonstrate changes in convective heat transfer in the pre-atherosclerotic regions, we compared intravascular voltage (IVV) between the NZW rabbits fed on high-fat/cholesterol (HD) and normal standard diets (ND). The MEMS thermal sensors detected significant changes in mean output voltage (Vave) and its time-varying component (δV/δt) in the pre-atherosclerotic regions that harbored active lipids and macrophages. Hence, the heat transfer strategy provides an entry point to link shear stress with pre-atherosclerotic regions.
Experimental Designs and Methods
Operation of the MEMS Thermal Sensors
The operational principle and fabrication process of the MEMS thermal sensors were previously described.36 Briefly, titanium (Ti) and platinum (Pt) were deposited on silicon wafers and patterned to form the sensing elements. The diameter of the coaxial wire was 0.4 mm and the sensing element was measured at 2 μm in width and 80 μm in span wise length (Fig. 1). The biocompatible epoxy (EPO-TEK 301: Epoxy Technology, MA) was used to anchor the sensor to the coaxial wire. Constant temperature (CT) mode was used to acquire real-time voltage signals.28 The CT circuit contained a feedback loop to maintain voltage across the sensing element, and was monitored by a LabVIEW-based data acquisition system at a sampling rate of 800 Hz.2,36 Calibration was performed in a 2-D flow channel for individual sensors and a relationship between heat loss and changes in output voltage over a range of flow rates was derived at 37.8 °C.3 The viscosity of the blood as a function of shear rate was obtained using a viscometer (Brookfield Engineering, Newhall, CA). Next, the MEMS sensors were deployed into the rabbit aortas for IVV measurements. Signal processing, wavelet decomposition, and low-pass filters were applied to minimize the background noise.31 The average signal-to-noise ratio (SNR) was approximately 4 in the rabbit aorta.
Protocol for NZW Rabbits
Six male rabbits (mean body weight 2.52 ± 0.30 kg) were acquired from a local breeder (Irish Farms, Norco, CA) and maintained by the Heart Institute of the Good Samaritan Hospital in Los Angeles. Prior to catheterization, the rabbits were anesthetized with an intramuscular injection of 50 mg/kg ketamine (Fort Dodge Laboratories, Inc) combined with Xylazine 10 mg/kg, with additional anesthesia given as required. A 23 gauge hypodermic needle and a 26 gauge guide wire were introduced into the left femoral artery via a cut-down. A rabbit femoral catheter (0.023″ID × 0.038″OD) was passed through the left femoral artery. The individual animals were heparinized (100 U/kg) prior to catheterization.
After IVV measurements, en face aortas were isolated for gross pathology, followed by immersion in 4% paraformaldehyde for 24 h. Vascular rings were frozen for histological analysis. Serial 5-μm cryosections of rabbit atheromas were cut (Sakura Finetek, Torrance, CA). Regions corresponding to the 3-D fluoroscope images were isolated, embedded, and sectioned. Next, immunostaining was performed with standard techniques in frozen vascular tissue using biotinylated secondary antibodies and peroxidase staining. Diaminobenzidine (DAB) was used as a chromogen and the sections were counterstained with hematoxylin for visualization of intima, media, smooth muscle cells, and adventitia. Atheromas, macrophages, or foam cells, were identified (Leica, DM LB2 microscope, Germany) and captured with a CCD digital camera (Spot RT-KE, Diagnostic Instruments, MI). Foam cells will be identified by Sudan black stain, lipids with oil-red-O,13,35 macrophages with anti-CD68 antibody, and oxLDL with mAb4E6.14
Three dimensional CFD code was developed for non-Newtonian fluid to simulate shear stress in the eccentric stenotic model.3 The average diameter of the rabbit abdominal aortas, Dave = 2.4 mm, was measured during catheterization. The non-Newtonian blood flow was simulated using the 3-D Navier–Stokes equations. The governing equations, including mass and momentum equations, were solved in FLUENT for laminar, incompressible, and non-Newtonian flow. The CFD solutions for the 3-D eccentric stenotic model in the presence and absence of catheter were previously reported.4 Briefly, the CFD model was constructed with 2, 410, 112 cells, which were primarily the tetrahedral elements. Fine elements, especially immediately adjacent to the tube wall and in the vicinity of the stenosis, were constructed to obtain sufficient information to characterize the large fluid velocity gradients near the wall. This approach guaranteed the mesh-independence of the numerical results. The simulations were performed under three steady state flow rates, namely, 50, 100, and 150 mL/min. The second order formulation of the solver was applied for the spatial discretization. The pressure–velocity coupling is based on the SIMPLEC technique.
All values were expressed as means ± SD. For statistical comparisons of IVV measurements and temporal gradients, we used a paired t test with values of p < 0.05 considered as significant. Comparisons of multiple mean values were performed by one-way analysis of variance (ANOVA), and statistical significance among multiple groups was determined using the Tukey procedure.
CFD Simulation of the 3-D Eccentric Stenotic Model
MEMS Sensors to Assess Convective Heat Transfer
Changes in IVV
The innovation of this study lies in the convective heat transfer strategy to detect changes in output voltage signals in the non-obstructive, albeit inflammatory, and otherwise considered normal TIMI flow regions. By virtue of its high temporal (7.5 kHz) and spatial resolution (100 μm), the MEMS thermal sensors enabled the detection of stenosis in terms of peak IVV in the throat of the stenosis and nadir in the region of flow reversal (Fig. 2).4 CFD codes further corroborated the peak and nadir in IVV measurements (Fig. 2c).4 Hence, convective heat transfer strategy detects changes in convective heat transfer in the pre-atherosclerotic regions in the fat-fed NZW rabbit model.
Real-time detection of changing flow profiles in the pre-atherosclerotic lesions remains a challenge. We hereby established a heat transfer strategy to detect flow reversal by deploying MEMS thermal sensors into the aorta of fat-fed NZW rabbits. Our findings revealed distinct IVV signals (Vave) accompanied by various temporal gradients (δV/δt) between lesion free and pre-atherosclerotic regions. In the distal aortic arch where flow reversal was prevalent in the pre-atherosclerotic regions, IVV measurements were significantly elevated. Despite geometric variations in the distal aortic arch, thoracic, and abdominal aortas, the IVV magnitudes remained nearly identical in the ND group (without pre-atherosclerotic lesions). Despite differences between ND and HD diets, the IVV magnitudes and δV/δt were similar in the abdominal aortas where atherosclerotic lesions were absent. Hence, the heat transfer strategy offers an entry point to assess micro-hemodynamics in the pre-atherosclerotic regions.
Heat transfer principle has been applied to study flow separation in response to low- vs. high-grade stenosis. Separation of shear layers from the arterial wall leads to a significant loss of viscous force (or a significant decrease in wall shear stress).4 Flow separation also develops in the arterial curvature and branching points.6,22 Helical flow developed lateral to the celiac artery and to the superior mesenteric ostia, accompanied by flow separation on the dorsal wall opposite to the superior mesenteric artery.5 Flow separation and flow recirculation engender disturbed flow or secondary flow, promoting recruitment of inflammatory cells.25 Our previous CFD codes provided insight into the pulsatile flow profiles and viscous secondary flow patterns in the branching points of rabbit abdominal aorta.5 We also showed the feasibility of using the MEMS thermal sensors to characterize regions of eccentric stenosis in an in vitro model.4 The current study demonstrated the capability of the MEMS thermal sensors to detect the pre-atherosclerotic regions by the means of convective heat transfer.
Magnetic resonance imaging20 and computational simulation29 have been used to characterize helical and retrograde secondary flow patterns in the aortic arch and branching points. Both methods revealed secondary motion,29 and flow reversals in the aortic arch and distal arch where turbulence remains localized, transient, and low in intensity.20 Complementary to these studies were the numerical and experimental data supporting the notion that secondary flow promotes heat transfer from the arterial wall to the flow field.8,32 In this context, our convective heat transfer strategy established the feasibility to detect time- and spatial-varying flow fields in the pre-atherosclerotic regions otherwise considered clinically silent during the conventional catheterization.
The MEMS thermal sensors operate on the convective heat transfer principle. Heat convection from the resistively heated element to the flowing fluid is measured as a function of the changes in voltage, from which shear stress can be inferred.23 As shown in Fig. 1, the voltage across the sensing element was significantly elevated in response to the secondary vortex motion that generated high fluid velocities and rapid heat dissipation from the surrounding fluid.8,32 The combination of a decrease in the sensor temperature and an increase in sensor resistance, coupled with an increase in power consumption, led to an increase in the voltage across the sensing element. Overall, detection and characterization of atherosclerotic regions are of utmost importance in the management of patients with suspected unstable atherosclerotic lesions. The heat transfer strategy hereby holds promises to assess changes in intravascular voltage in the non-obstructive, albeit pro-inflammatory regions during angiograms.
The authors would like to express their gratitude to Dr. Herbert J. Meiselman from Physiology and Biophysics for his assistance with viscosity measurement. The authors are grateful for the technical support from Sharon L. Hale and Dr. Robert A. Kloner at the Heart Institute of Good Samaritan Hospital. This work was supported by AHA Pre-Doctoral Fellowship 0615063Y (L. A.), NHLBI HL 83015 (T. K. H.), and NHLBI HL091302 (T. K. H.).
- 9.Chatzizisis, Y. S., M. Jonas, A. U. Coskun, et al. Prediction of the localization of high-risk coronary atherosclerotic plaques on the basis of low endothelial shear stress—an intravascular ultrasound and histopathology natural history study. Circulation 117(8):993–1002, 2008.PubMedCrossRefGoogle Scholar
- 11.Fung, Y. C. Biomechanics: Circulation (2nd ed.). New York: Springer, 1997.Google Scholar
- 12.Haritonnidis, J. H. The measurement of wall shear stress. In: Advances in Fluid Mechanics measurements. New York: Springer-Verlag, 1989, pp. 229–236.Google Scholar
- 18.Imoto, K., T. Hiro, T. Fujii, et al. Longitudinal structural determinants of atherosclerotic plaque vulnerability: a computational analysis of stress distribution using vessel models and three-dimensional intravascular ultrasound imaging. J. Am. Coll. Cardiol. 46(8):1507–1515, 2005.PubMedCrossRefGoogle Scholar
- 23.Liu, C., Y. C. Tai, J. B. Huang, et al. Surface Micromachined Thermal Shear Stress Sensor. The First ASME Symposium on Application of Micro-Fabrication to Fluid Mechanics, pp. 9–15, 1994.Google Scholar
- 32.Tatsumi, K., H. Iwai, K. Inaoka, et al. Secondary flow effect to heat transfer of a duct with discrete rib turbulators. Heat Transf. Gas Turbine Syst. 934:481–488, 2001.Google Scholar
- 34.Topper, J. N., J. Cai, D. Falb, et al. Identification of vascular endothelial genes differentially responsive to fluid mechanical stimuli: cyclooxygenase-2, manganese superoxide dismutase, and endothelial cell nitric oxide synthase are selectively up-regulated by steady laminar shear stress. Proc. Natl Acad. Sci. 93(19):10417–10422, 1996.PubMedCrossRefGoogle Scholar
- 36.Yu, H., L. Ai, M. Rouhanizadeh, D. Patel, E. S. Kim, and T. K. Hsiai. Flexible polymer sensors for in vivo intravascular shear stress analysis. IEEE/ASME J. MEMS 17(5):1178–1186, 2008.Google Scholar