Paper Submission & Registration
8th Dutch Bio-Medical Engineering Conference
Go-previous
15:20   Heart - IV
Chair: Peter Veltink
15:20
15 mins
Single- and multiple- tracking location(s) shear wave elastography for viscosity mapping by system identification
Xufei Chen, Rogier R. Wildeboer, Alexander F. Kolen, Ruud J.G. van Sloun, Massimo Mischi
Abstract: Viscosity has been shown to be a tissue property of diagnostic value that plays an important role in the propagation of shear waves. Recently, system identification (SI) techniques have been proposed for the estimation of tissue viscoelasticity maps. Here, we compare the performance of SI-based viscosity imaging by multiple tracking location (MTL) shear wave elastography (SWE) and by single tracking location (STL) SWE. The latter reduces speckle bias at the cost of using multiple push pulses to map the entire field of view.
15:35
15 mins
Improving image quality and functional measurements of echocardiography using multi-perspective ultrasound
Marloes Sjoerdsma, Sjoerd Bouwmeester, Patrick Houthuizen, Frans van de Vosse, Richard Lopata
Abstract: In the clinic, ultrasound is generally the modality of choice due to its high temporal and spatial resolution, availability, and patient safety. Moreover, 3D ultrasound captures the heart’s complex shape and contraction pattern more accurately. However, echocardiography has a limited field-of-view, poor image contrast, and low lateral resolution. Furthermore, echocardiography is prone to clutter artifacts in the near-field. Consequently, the distinction between the cardiac walls and the blood-pool is faded. These drawbacks accumulate in 3D, due to lower spatial and temporal resolution. To overcome these limitations, a widely applicable method to merge echocardiograms from different locations (i.e. multi-perspective ultrasound) was developed and validated to improve the image quality, enhance functional measurements, and reduce clutter in the near-field, whilst preserving the ultrasound speckle required for strain imaging. 3D Echocardiograms were obtained at different positions on the chest in 10 healthy volunteers. Additionally, 3D ultrasound data were acquired in 20 patients for the clinical validation of the clutter reduction algorithm. In order to include all anatomical structures whilst reducing noise as much as possible in the merged resulting image, the echocardiograms were decomposed into a subset of images, each containing structures of a different size range. The anatomical structures were preserved by taking the maximum intensity of the images containing the bigger structures. For noise reduction, only the small structures in the axial direction were included, which enable speckle tracking needed for cardiac strain imaging. The method proposed increased the field-of-view by 5-25%. Moreover, it improved the lateral resolution by a factor three, and the images contained 8-28% more anatomical structures in the overlapping area. The filter reduced clutter in all datasets, whilst preserving the cardiac walls. Additionally, the image contrast-to-noise and signal-to-noise ratio increased by 4.2 and 3.6 dB, on average, respectively. Finally, global longitudinal strain measurements improved in 6 out of 10 subjects. In conclusion, echocardiographic image quality and functional measurements were improved using multi-perspective ultrasound, which includes a near-field clutter reduction algorithm.
15:50
15 mins
Shear stress related plaque growth of lipid-rich plaques in human coronary arteries
Eline M.J. Hartman, Giuseppe De Nisco, Annette M. Kok, Fay M.A. Nous, Alya Hoogendoorn, Frits Mastik, Suze-Anne Korteland, Frank J.H. Gijsen, Anton F.W. van der Steen, Ricardo Budde, Joost Daemen, Jolanda J. Wentzel
Abstract: Background Low wall shear stress (TAWSS) plays a vital role in the development and progression of atherosclerotic plaques. Little is known on the differences in plaque progression between lipid-rich (LRP) versus non-lipid rich plaque as assessed by near-infrared spectroscopy (NIRS) and on the interplay with TAWSS. This study aimed to investigate the combined effect of WSS and lipid content on plaque progression. Methods Serial imaging with near-infrared spectroscopy-intravascular ultrasound (NIRS-IVUS) of at least one non-culprit vessel was performed in patients with an acute coronary syndrome at baseline and 1-year follow-up. IVUS segmented lumen information was fused with a CT-derived vessel centerline to create a 3D model (Figure 1). Time resolved WSS (100 time steps per cardiac cycle) was computed by applying computational fluid dynamics (ANSYS, Fluent (v17.1) to the models using invasive Doppler flow measurements as outflow boundary conditions at the side branches. Each artery was divided into 1.5mm/45° sectors, which were classified as either lipid positive or negative based on the NIRS-signal. The average TAWSS per sector was divided into vessel-specific tertiles. By using mixed ANCOVA models, plaque growth based on wall thickness (WT) change of the different subgroups was assessed. Results For 39 vessels, complete imaging was available. NIRS positive sectors showed more plaque progression than NIRS negative sectors independently from baseline WT (Figure 2, 0.06 mm 95%CI:0.034-0.075 vs 0.0 mm 95%CI 0.013-0.019, p<0.001). Sectors exposed to low WSS displayed more plaque progression than sectors exposed to mid or high WSS (0.034 mm 95%CI:0.028-0.040; 0.009 mm 95%CI:0.004-0.0015; 0.011 mm 95%CI:0.005-0.018 p<0.001). Finally, combining NIRS and WSS showed an incremental effect on the plaque progression over time. Conclusion Lipid rich sectors exposed to low WSS demonstrated the highest plaque progression over time.
16:05
15 mins
The effect of electrode size on electrogram array properties
Bahareh Abdi Abdi, Mathijs van Schie, Richard Hendriks, Natasja de Groot
Abstract: Epicardial atrial electrograms (EGM) play an important role when studying the initiation and perpetuation of atrial fibrillation (AF) or guiding the cardiologist through ablative therapies. In clinical practice many different catheters/electrodes are used for recording EGMs. It is unknown how exactly the electrode size affects EGM morphology. However, these effects can critically influence AF treatment by physicians performing ablation. In this study, we use the electrode's transfer function to model and analyse the effect of electrode size on the properties of measured EGM. Initially, we simulate EGM arrays using a two-dimensional (2D) EGM model. For true data, however, we first estimate the trans-membrane currents from the measured EGMs for a 2D cell grid at a much higher resolution than the original electrode array and later use these values to model EGMs with different electrode sizes. With respect to the simulated data, we simulate EGM arrays for 2D tissues with different levels of heterogeneity in conduction and stimulation pattern to model the inhomogeneous wave propagation during AF. The results show that increasing the electrode size increases the error in local activation time (LAT) estimation with a decrease in the length of observed block lines in the recording area. Moreover, visual inspection shows that fewer irregularities and wavelets can be observed using larger electrode sizes. The increase in electrode size also increases the low voltage areas in the tissue while decreasing the maximum steepness of the activities and the number of detected deflections. The effect is more pronounced for simulated tissues with higher level of heterogeneities in the conduction pattern. The same measures are also computed for true EGM arrays recorded from Bachman bundle of 10 patients during AF. The results show similar results as in the simulated data. Increasing the electrode size thus affects the properties of recorded EGMs including increasing the error in LAT estimation, decreasing the length of block lines, increasing the low voltage areas, decreasing the steepness of the activities and the number of deflections. This is of great importance that physicians consider these effects while performing any analysis on the EGMs or while determining a treatment approach.


end %-->