6th Dutch Bio-Medical Engineering Conference
26 & 27 January 2017, Egmond aan Zee, The Netherlands
13:00   Ultrasound Imaging II
Chair: Chris de Korte
15 mins
Stein Fekkes, Anne Saris, Maartje Nillesen, Hendrik Hansen, Chris de Korte
Abstract: Development of atherosclerosis in the carotid artery (CA) elevates the risk for severe stenosis obstructing the main cerebral blood flow supply due to geometrical changes of the vessel wall. It also initiates compositional changes due to the development of lipid pools and calcifications in the vessel wall, which could lead to rupture prone plaques. These pathological processes can be asymptomatic for a long time until the initiation of a catastrophic event such as a stroke or transient ischemic attack. This is why characterization of the stage of atherosclerosis is important to guide pharmaceutical treatment or perform an invasive intervention removing suspected critical parts of the CA. To aid in the anatomical and functional assessment of the CA, ultrasound B-mode imaging overlaid with either flow or strain information has been widely adopted. Recently ultrafast ultrasound imaging has become available. Ultrafast imaging has enabled 2D vector velocity blood flow imaging as well as improved compound strain imaging. Delivering the blood velocity and strain information simultaneously might aid the clinician in establishing a fast and reliable judgment about the overall geometrical and functional progression of the disease stage. This study compares the performance of a high frequency (> 20MHz) transducer with a clinically utilized (9 MHz) transducer in delivering this information. A realistic patient specific phantom made of polyvinyl alcohol was constructed of the CA-bifurcation. The vessel wall was embedded in a surrounding environment simulating the adventitia with extended surrounding tissue. A realistic pulsatile flow of 60 bpm was imposed on the phantom using blood mimicking fluid, with a mean of 0.6 L min-1 and a peak flow of 1.4 L min-1 resulting in a fluid pressure of 150 over 100 mmHg. Continuous plane wave ultrasound acquisitions at 4000 fps were performed using a Verasonics Vantage research ultrasound system connected with a Visualsonics MS250 (fc = 21 MHz, pitch = 88 µm, 256 elements) and a Philips L12-5 transducer (fc = 9 MHz, pitch = 195 µm, 256 elements). A translation setup and a side-by-side probe mold ensured perfectly aligned imaging planes enabling exact field of view registration and comparison. A cross-correlation based coarse-to-fine method scaled with the center frequency was applied on the beamformed data to determine the axial and lateral strain of the vessel wall and, after clutter removal, the velocity magnitude and direction of the blood mimicking fluid. Finally, flow and strain estimates were merged at an effective frame rate of 100Hz covering the full pressure cycle. Results show that the combination of plane wave and high frequency imaging is feasible to obtain both flow and strain information. Improved axial and lateral resolution seems beneficial for both strain and flow estimation. The application of high frequency ultrafast imaging is however limited to structures at shallow depths (<1.5 cm).
15 mins
Maysam Shabanimotlagh, Shreyas Raghunathan, Deep Bera, Zhao Chen, Chao Chen, Verya Daeichin, Michiel Pertijs, Hans Bosch, Nico de Jong, Martin Verweij
Abstract: Echocardiography is a portable, safe, and low-cost imaging technique for accurate assessment of the heart. In transesophageal echocardiography (TEE) the esophagus is utilized as the imaging window to examine the cardiac anatomy and function. In conventional TEE probes, a one-dimensional (1D) ultrasound array is employed to obtain two-dimensional (2D) cross-sectional images of the heart. Since cardiac morphology, leakage of valves and function of the outflow tracts are all three-dimensional (3D) phenomena, it is beneficial to interpret them from 3D images. Therefore, there is high clinical demand for matrix TEE probes that are capable of providing real-time volumetric images [1]. Several matrix arrays (Philips X7-2t, Siemens V5M TEE, General Electric 6VTD) have been developed for this purpose, however all of them are large in size (~10 cm3) and uncomfortable to use on non-anesthetized patients [2]. We aim to develop a matrix TEE probe with a small head volume (<1 cm3), which is suitable for long term monitoring of cardiac system on adults and in babies. We have developed a prototype of a small matrix TEE probe, which consists of a piezoelectric matrix transducer directly mounted on an Application Specific Integrated Circuit (ASIC). The ASIC performs the task of micro-beamforming, signal amplification and efficient data reduction. The piezoelectric matrix array consist of a 32×32 PZT elements with a pitch of 150 μm × 150 μm. The transmit aperture consists of 8×8 elements at the centre of the array, which are directly wired out to the ultrasound system. The remaining 864 elements are used in receive and are organized in 96 sub-arrays of 3×3 elements to reduce the cable count with a factor of 9. The signals from the individual elements in a sub-array are combined to a single output signal using a micro-beamformer on the ASIC. The micro-beamformer allows pre-steering of 0◦, ±17◦, and ±37◦ angles in both lateral and elevation directions. By recording datasets for different pre-steering angles, and by processing and combining them, a large volume image can be constructed. Acoustic performance of the prototype is evaluated in a water tank. The transmit transfer function of a single element is measured by applying a 20 cycle sinusoidal voltage, sweeping from 3 to 8 MHz with steps of 50 kHz. The output pressure is recorded by a calibrated hydrophone. It is found that the transducer has a central frequency of 5 MHz, a bandwidth of 40% and a transmit efficiency of 6.4 kPa/V (at 51 mm). To characterize the micro-beamforming function, three delay angles of 0◦, 17◦ and 37◦ were programmed. While transmitting with a well-defined external source, the output voltage from a sub-group was recorded from -50◦ to +50◦ degrees. We observe that the theoretical values of the beam profile agree well with the measurement results, especially with regard to the position of the grating lobes and side lobes.
15 mins
Umit Arabul, Maarten Heres, Marcel Rutten, Marc van Sambeek, Frans van de Vosse, Richard Lopata
Abstract: Vulnerability assessment of carotid plaques is vital to prevent stroke and stroke-related mortality. The composition of a carotid plaque plays an important role to assess its stability. The optical wavelength specific response of photoacoustic (PA) imaging is currently explored to aid in the diagnosis of atherosclerosis in carotid arteries. Using multiple wavelengths, PA has the potential to reveal vital morphological information in plaques such as intraplaque hemorrhages, lipid pools, and the fibrous cap. In this study, we used PA and plane wave ultrasound (PUS) hybrid-imaging to reveal the compositions of endarterectomy samples. Secondly, we investigated the effects of multi-angular spatial compounding based on SNR comparison of ex vivo carotid plaques. Plaque samples (N = 7) were obtained from a local hospital after endarterectomy. The samples were mounted in an experiment setup and imaged (cross-sectional) with an integrated handheld PA probe, consisting of an integrated diode laser (Ep = 1 mJ, tp = 130 ns, λ = 808 nm, QUANTEL, FR) and a linear array transducer (fc = 7.5 MHz, ESAOTE, NL). During the acquisition, the probe position was swept longitudinally to obtain the complete 3D volume of the plaque sample. Next, the sample was rotated by 10° steps and measurements were repeated for 36 angles. Data were spatially compounded for complete 360° rotation and for the in vivo available range ( -30°, +30°) of rotation. Finally, compounding using the rotation range feasible in vivo ( -30°, +30°) was compared to the full tomographic scan. Areas of high absorption in the 3D datasets were identified and compared to histological data of the plaques. Photoacoustic images of plaque samples revealed morphological information that is not visible in PUS images. Data in six out of seven endarterectomy samples revealed the presence of intraplaque hemorrhages and demonstrated that PA imaging of carotid plaques at this wavelength is capable of detecting intraplaque hemorrhages. Moreover, the compounding results showed that spatial compounding elevated the SNR 5 by dB in plaque samples. Additionally, the compounding results for the limited range of rotation as in vivo ( -30°, +30°) showed 7.02 ± 2.73 dB SNR enhancement. Compounded PA data of only seven angles (with 10° steps) are comparable to full tomography results and could provide a sufficient SNR for in vivo PA imaging of plaques, however, only for the proximal wall. Due to the non-invasive nature of photoacoustic imaging, this ex vivo study may elucidate the pre-clinical studies towards the in vivo, non-invasive vulnerability assessment of the carotid plaques.
15 mins
Louis Fixsen, Niels Petterson, Frans van de Vosse, Marcel Rutten, Richard Lopata
Abstract: Patients with end-stage heart failure are increasingly being treated with left ventricular assist devices (LVADs). Much of the interplay between the device and the (functioning sections of the) heart is still unknown however. Cardiac strain imaging is a powerful imaging technique that could enable the quantification of clinically significant parameters such as strain and strain rate in the hearts of patients with mechanical support. Strain imaging involves tracking the motion and deformation of tissue in an ultrasound (US) image. The reproducibility and accuracy of such measurements remain issues¬. Typical methods for validation such as phantoms and simulation lack realism, whilst in vivo animal studies have ethical issues, especially at the scale needed for validation purposes. In this study, use of an isolated beating porcine heart platform for the assessment of US strain imaging in LVAD supported hearts was investigated. 2-D US data of the short-axis view of the left ventricle were acquired at increasing LVAD pump speeds. The ex vivo porcine hearts were implanted with Thoratec HeartMate II (n = 2) and MicroMed DeBakey (n = 2) LVADs, with the inflow at the apex and the outflow at the aortic tube. The hearts were then attached to a mock circulation loop and re-perfused with warm oxygenated blood, before being resuscitated and paced at a stable rate close to 120 beats-per-minute. The beating hearts were submerged in water to facilitate US imaging. Radiofrequency (RF) data were acquired using a modified Esaote MyLab 70 XVG ultrasound system and curved array transducer. Data were acquired at pump speeds between 0 (with outflow clamped shut) and 10.5 thousand revolutions per minute (krpm). Cross-sections were imaged at three points along the length of the ventricle, from apex, above the LVAD inflow, to mitral valve. Data were manually segmented and displacement fields for each frame were calculated and tracked at each point of the mesh [1]. Finally, the displacements were converted into radial and circumferential strain between end-diastole and end-systole, using a least-squares strain estimator. Good agreement was found in radial and circumferential strain between individual heart cycles with the pump outflow clamped shut (ICC’s of 0.93 and 0.92 respectively) and at higher pump speeds (ICC’s of 0.96 and 0.83). Notable also is the change in pattern of circumferential strain as the pump speed was increased past 8 krpm. Future work will involve investigating the change in strain pattern further through parameters such as strain rate.
15 mins
Khalid Daoudi, Martijn Hoogenboom, Dylan Eikelenboom, Gosse Adema, Jurgen Fütterer, Chris de Korte
Abstract: Mechanical HIFU ablation can play a major role in providing therapeutic means for oncology. While thermal ablation has made its way to the clinic, mechanical ablation stumbles still in preclinical investigations due to the lack of understanding of its pathological and immunological impact. Several mechanical effects can be created by HIFU ablation depending on the type of treatment and we investigated in this work boiling histotripsy [1]. This relatively new technique uses extreme high US pressure to destroy and pulverize the targeted tissue. Consequently, a better understanding of the impact on the tumor immunology and pathology and long term implications is required. To address this lack of comprehension, we need to study in-vivo and non-invasively the effect in different aspects and an adequate real-time imaging modality would be beneficial for a better understanding. In this work, we investigate the possibility to use multimodal imaging to detect different effects caused by the boiling histotripsy ablation. For the purpose of the study, two C57Bl/6n wild type mice were prepared and subcutaneously injected with neuroblastoma tumor cells in the right femur and both were treated after 4 to 5 weeks with high-pressure amplitudes. The ablation takes place in a 7T wide bore animal MR scanner where the ablated region is monitored during the treatment. A 3MHz HIFU system, focal spot size 0.5x0.5x2.0mm was used for the treatment. For boiling histotripsy, 5ms pulses with an acoustic output power of 230 Watt, a pulse repetition frequency of 1Hz and 200 pulses per focal spot were used. We utilized Photo-Acoustic (PA) imaging with co-registered high frequency ultrasound (HFUS) to non-invasively image the tumor before and after HIFU ablation. After the imaging, the tumor was resected for further assessment and evaluation of the ablated region using histopathology. Photoacoustic technique, mainly sensitive to blood, revealed the presence of large absorbing structure composed of deoxygenated blood while we mainly detected oxygenated vessels before treatment. The deoxygenation is caused by the presence of red blood cells outside the vessels due to their rupture or destruction as a result of the ablation. Ultrasound images show hypoechoic region in the treated area revealing the complete pulverization of the tumor cells whereas Doppler imaging revealed the destruction of blood vessel network after the ablation. The findings were supported by histopathology.
15 mins
Gert Weijers, Geert Wanten, Johan Thijssen, Marinette van der Graaf, Chris de Korte
Abstract: Patients on home parenteral nutrition are at risk for developing liver dysfunction, which is partly due to the accumulation of lipids in the liver (steatosis) and may progress to end-stage liver disease with certain liver failure. Therefore, a timely diagnosis with easy access to repeated assessment of the degree of liver steatosis is of great importance. A pilot study was performed in 14 patients on long-term home parenteral nutrition using the computer-aided ultrasound (CAUS) method [1]. Ultrasound B-mode images (n=5) were acquired using a phased array transducer. All patients were subjected to proton magnetic resonance spectroscopy measurement of liver fat content for reference. Gain compensation, large hepatic blood vessels and bile ducts segmentation, and attenuation correction are performed automatically in CAUS. The CAUS software produces histogram and texture parameters, both average ROI values and as a function of depth. All parameters are expressed relatively to the tissue mimicking phantom (TMP) used for equipment and imaging preset calibration. The most significant correlating parameters to the MRS fat content were found to be the residual attenuation coefficient (RAC, R=0.87, p<0.001), the depth dependence of the lateral speckle size (LATs, R=0.77, p=0.001), and the mean echo level (MU, R=0.71, p=0.007). Multiple linear regression analysis revealed a R2 of 0.85 based on the RAC parameter only. Findings corroborate with previously performed study on cows, and thus indicate the potential usefulness of CAUS for staging of human hepatic steatosis. However, extension of the study is warrant for prognostic values estimation. REFERENCES [1] G. Weijers, G. Wanten, J. M. Thijssen, M. van der Graaf, and C. L. de Korte, “Quantitative Ultrasound for Staging of Hepatic Steatosis in Patients on Home Parenteral Nutrition Validated with Magnetic Resonance Spectroscopy: A Feasibility Study,” Ultrasound Med Biol, vol. 42, no. 3, pp. 637-44, Mar, 2016.