5th Dutch Bio-Medical Engineering Conference 2015
22-23 January 2015, Egmond aan Zee, The Netherlands
10:40   Cardiovascular Mechanics I
15 mins
Stefan Sanders, Frans van de Vosse, Marcel Rutten
Abstract: Vulnerable plaques in carotid arteries are potentially lethal or at least cause high risk of stroke when they rupture. Vulnerability research has focused on estimating geometric properties such as cap thickness and fluid dynamics, evaluating wall shear stress. Recently, the detection of plaque properties using combined imaging/numerical methods is gaining popularity. Three-dimensional methods using MRI data and fully 3D numerical models have shown the possibilities of assessing the mechanical environment of plaques, although the handling of heterogeneity is still a problem due to spatial resolution issues. Higher resolution imaging methods like ultrasound cannot make a sharp distinction between plaque components. Studies on a microscopic scale, including indentation tests on human plaque samples have elucidated the highly heterogenic properties of pathological tissue. Such tests, however, load the material in a direction which does not coincide with physiological loading directions, and thus may yield results that cannot be translated easily to clinical applications. In this study, we model a slice of artery or arterial plaque as a 2D object, and load it in a physiological way. Slices (0.3mm thickness) were cut from a fresh porcine carotid artery, obtained from the slaughterhouse. The slices were mounted between two horizontal microscope slides, 0.25mm apart. Through a small hole in the bottom slide, transparent paraffin oil was injected into the lumen of the artery, thus pressurizing the vascular tissue from the luminal side, and enabling the tissue to move frictionless. Inflow pressure was measured using a standard clinical pressure sensor. Motion of the sample was measured using a high speed video camera. The images were processed using Matlab to render diameter and wall thickness as a function of pressure. To assess mechanical properties, the experiment was modelled in FEBio finite element code (v, using quadratic hexahedral elements. To start, the vascular material was modelled as being neo-Hookean. The experiment was modelled and experimental and simulation results compared. The method proposed gives good results for the mechanical properties of vascular tissue, when compared with standard testing techniques. Friction between glass coverslips and sample material proved to be not a large problem, which was anticipated. The effects of friction were negligible for the properties estimation. As such, the method can be applied to assess properties of small samples of vessels exhibiting axial heterogeneity. The main advantage is the physiological loading condition imposed in a quasi 2D setting, and the small amount of tissue required to do the test. In future applications, heterogeneous properties, like in atherosclerotic plaque material, may be assessed as well, using vital staining techniques to discern different tissue components without compromising their mechanical behaviour.
15 mins
Janina Schwarz, Elco Oost, Pepijn van Horssen, Monique van Lier, Erik Bakker, Jeroen van den Wijngaard, Jos Spaan, Maria Siebes
Abstract: Background: Novel biomarkers developed for early-stage visualization of neovascularization with clinical imaging techniques require validation in ischemic animal models. Vascular adaptation to ischemia co-localized with fluorescently labelled biomarkers can be characterized in 3D with an imaging cryomicrotome [1]. Specimen preparation and vascular filling were optimized for detection and diameter assessment of collateral vessels in rat hind limbs. Methods: The hind limb vasculature in healthy (n=27) and ischemic (n=18) rats was filled with fluorescent cast material (Mercox, Batson’s #17) of varying viscosity to optimize penetration into small vessels. The frozen casts were sequentially sliced at 24 µm thickness and imaged, resulting in co-registered reflectance and fluorescence 3D image stacks. Raw data were processed to reduce the influence of fluorescent signals from underlying tissue, enhance tubular structures, extract vessel centrelines and identify vessels and collateral connections. Diameter assessment was calibrated with sutures and tubes of known diameter (20 – 510 µm). Results: An optimal viscosity range (25-30% dilution) was established for infusion into the abdominal aorta at physiological pressure in healthy rats. Unbalanced filling in ischemic rats with unilateral femoral ligation was overcome by initially clamping the healthy iliac artery. Measured diameters agreed well with actual sizes (R²=0.997). Collateral vessels could be visualized in ischemic and in non-ischemic hind limbs. As a proof of principle, vasculature in the legs of healthy rats was segmented; collateral connections were automatically detected. Conclusion: Serial cryo-sectioning and epifluorescence imaging provides a powerful method to study mechanisms of neovascularization by 3D quantification of collateral adaptation in vascular networks.
15 mins
Merih Cibis, Wouter Potters, Mariana Selwaness, Frank Gijsen, Oscar Franco, Andres Lorza, Aad van der Lugt, Aart Nederveen, Jolanda Wentzel
Abstract: Introduction Atherosclerotic plaque initiation is associated with the level of wall shear stress (WSS) (1) which is generally calculated by computational fluid dynamics (CFD) and more recently by using phase-contrast MRI (PC-MRI).Although MRI-based WSS (WSSMRI) shows generally good qualitative agreement with CFD-based WSS (WSSCFD), the magnitude and the direction of WSS may differ in these two methods (2). Currently a debate is going on, whether the observed relationships between wall thickness and WSS are realistic, since not many studies show a point-to-point relationship (3). It is therefore of vital importance to know if the observed differences in WSS would affect the conclusions regarding plaque initiation and progression. In this study, our objective was to compare the associations between WSSMRI and WSSCFD and arterial wall characteristics. Methods 16 elderly subjects with plaques in their left carotid arteries were selected from the Rotterdam study (cohort study in a suburb of Rotterdam). These subjects underwent an MRI scan and the left carotid arteries of the subjects were segmented from the MR images and used for steady state CFD simulations applying MRI based inflow velocity profiles as boundary conditions. The lumen area, outer wall area, wall thickness (WT), WSSMRI and WSSCFD were calculated. WSS was correlated with WT in ICA and CCA. Results WSSMRI (0.48±0.26Pa) was lower than WSSCFD (0.82±0.54Pa) in the ICA at baseline, but they showed qualitatively overlap in their patterns. In the CCA, the WSS values based on CFD and MRI were almost equal (WSSMRI: 0.50±0.26Pa and WSSCFD: 0.54±0.65Pa) but the WSS patterns were different caused by the steady state simulations which eliminate transient inertial effects. In ICA, WT was significantly inversely related with both WSSMRI (p<0.001) and WSSCFD (p=0.007). Interestingly, in CCA, this was only the case for WT-WSSMRI relationship (p=0.005). Conclusions The magnitude of WSS differs when obtained with CFD and PC-MRI in ICA, but the inverse relation between arterial wall thickness and WSS can be captured by both methods. Whereas in CCA, the mean WSS was almost equal, but only using PC-MRI for WSS assessment resulted in an inverse relation between wall thickness and WSS. Therefore, WSS in CCA must be interpreted more carefully. REFERENCES [1] Davignon, J, Ganz, P. Role of Endothelial Dysfunction in Atherosclerosis. Circulation 2004 Jun 15; 109: III27-32. [2] Cibis, M, Potters, WV, Gijsen, FJH, Marquering, H, vanBavel, E,van der Steen, AFW, Nederveen, AJ, Wentzel JJ. Wall shear stress calculations based on 3D cine phase contrast MRI and computational fluid dynamics: a comparison study in healthy carotid arteries. NMR Biomed. 2014 Jul; 27(7):826-34. [3] Peiffer V, Sherwin SJ, Weinberg PD. Does low and oscillatory wall shear stress correlate spatially with early atherosclerosis? A systematic review. Cardiovasc Res. 2013 Jul; 99(2); 242-250.
15 mins
Hendrik Hansen, Stein Fekkes, Anne Saris, Gerard Pasterkamp, Michiel Bots, Gert Jan de Borst, Frans Moll, Maartje Nillesen, Chris de Korte
Abstract: Plaque rupture in the carotid artery is the source of most strokes and transient ischemic attacks. Pathology studies have shown that plaques which contain a lipid-rich pool covered by a thin fibrous cap have a higher change to rupture than plaques which mainly contain fibrous tissue. Deformations of the arterial wall induced by the pulsating blood indirectly provide information on mechanical plaque composition. Over the past years we developed a technique, called compound ultrasound strain imaging, to estimate these deformations noninvasively. The compounding technique provides highly accurate estimates of the strain tensor by combining displacement estimates obtained by cross-correlation of raw radiofrequency (RF) ultrasound data acquired at three insonification angles. This study consists of two parts. In the first part the in vivo performance of the compounding technique to detect rupture prone (vulnerable) plaques was assessed in patients (n=34) with highly stenotic arteries (>70%). Strains were estimated before surgical excision of the plaque. Histological staining was applied to the excised plaques and features of vulnerability were scored by an expert blinded to the strain results. Receiver-operating curves were constructed to assess the performance of the compound strain imaging technique to detect these histological features of vulnerability. The technique successfully differentiated between plaques with a large lipid pool and fibrous plaques with a sensitivity, specificity, positive and negative predictive value of 85%, 71%, 81% and 77%, respectively. For the individual features of vulnerability, such as fibrous cap thickness, the technique was less conclusive. As second part of the study, the compounding method was applied on ultrasound data acquired using a different way of ultrasound transmission. Instead of conventional focused transmissions per image line, one image was recorded at a time by transmitting plane wave ultrasound. In this way, imaging frame rates were increased by a factor of 128, albeit at the cost of image quality, especially in the direction perpendicular to the ultrasound beam. However, angle misalignment errors during compounding were expected to be reduced, because of the faster acquisition rate. Our first strain results obtained with plane wave transmission, for simulations of an artery with a vulnerable plaque and experimental data of a two-layered vessel phantom, are of similar precision as the strain results obtained with focused ultrasound transmission. Thus, strain estimation at kilohertz frame rates is possible without loss of precision. This is an important finding, since it indicates that the technique might also be extended to 3-D. Combined with the in vivo results, this implies that vulnerable plaque detection in 3-D might be possible in future.
15 mins
Guy Warmerdam, Rik Vullings, Jan Bergmans, Guid Oei
Abstract: Valuable information about fetal wellbeing during labor could be obtained from spectral analysis of fetal heart rate variability. However, extracted fetal heart rate recordings are often contaminated by artefacts that disturb the spectral analysis. Although it is possible to correct for these artifacts, this correction itself will also affect the spectral analysis. In this study, a method is presented that estimates the error in spectral power that is caused by artifact correction. Clinicians can use the information about this error to assess the reliability of spectral analysis of artifact corrupted fetal heart rate recordings.
15 mins
Constantin Ungureanu, Ronald Aarts, Johan Arends
Abstract: Heart rate, respiratory rate, body temperature, oxygen saturation, and blood pressure are the main vital signs of the human body [1]. Monitoring of these parameters can provide information about the general health status of patients. In general, each vital parameter is measured with a different device. In this paper we present the details of a preclinical prototype for measurement of heart and respiratory rate plus oxygen saturation. This device is embedded in a glove that has an optical sensor to record the photopletysmographic (PPG) signal from the patient. The optical module situated on the palmar side of the glove uses a reflection mode pulse oximeter at 25 degrees incidence. Further, the PPG signal is recorded and transmitted wirelessly by a Shimmer sensor node situated on the dorsal side of the glove. On the same side a solar cell is located which is used to charge the battery of the glove system. The PPG signals are received by a laptop and processed in real time to extract the vital signs. To measure the vital signs, the nurse or the doctor simply touches the patient skin while using the glove. We evaluated the glove on 10 healthy volunteers. The respiratory and heart rate errors were under 4 breaths respective beats per minute when compared with reference. This device can be used in rural areas by doctors to evaluate the general health status of patients. In addition, the glove can communicate with smart phones where databases can be generated.