Catálogo de publicaciones - revistas

<jats:title>Abstract</jats:title> <jats:p>The assessment of mechanical stiffness is an essential diagnostic tool for investigating the biomechanical properties of biological tissues. Surface wave elastography (SWE) is an emerging technique to quantify elastic properties of tissues in clinical diagnosis. High-speed optical imaging combined with SWE has enormous potential in quantifying the elastic properties of tissues at microscale resolutions. In this study, we implement surface wave elastography using high-speed optical interferometry to characterize the elastic properties of tissue-mimicking phantoms and <jats:italic>ex-vivo</jats:italic> native caprine liver tissue by imaging the surface wave induced by an electromechanical actuator. The sinusoidal mechanical excitations ranging from 120 Hz to 1.2 kHz on the surface of tissues are captured using a high-speed camera with a frame rate of 4 kHz at micrometer resolutions. The surface wavefront reconstruction is performed using a phase-shifting algorithm and linear regression is used to calculate the surface wave velocity. The mechanical stiffness estimated from the optical system is compared with the results of mechanical compression testing measurements. The results from this multimodal platform combining optical interferometry and vibrational spectroscopy using SWE are highly promising towards a non-invasive or minimally invasive imaging for <jats:italic>in-vivo</jats:italic> and <jats:italic>ex-vivo</jats:italic> mechanical characterization of tissues with future clinical applications.</jats:p>

<jats:title>Abstract</jats:title> <jats:p> <jats:italic>Purpose</jats:italic>. This is a dosimetric study comparing stereotactic body radiotherapy (SBRT) plans of spine tumors using Brainlab Elements Spine planning module against Eclipse RapidArc plans. Dose conformity, dose gradient, dose fall-off, and patient-specific quality assurance (QA) metrics were evaluated. Methods:Twenty patients were immobilized in supine position using half Vac-Lok. A prescription dose of 16 Gy in a single fraction was planned for Varian TrueBeam. Conformal arc plans were generated with Pencil beam (PB), MonteCarlo (MC) in Elements, and RapidArc with Acuros XB algorithm in Eclipse using identical treatment geometry. <jats:italic>Result</jats:italic>s. Eclipse, Elements PB, and Elements MC generated dosimetrically conformal plans having Inverse Paddick Conformity Index (IPCI) &lt;1.3. All plans satisfied the dose constraints to target and OARs. Elements PB had a sharper gradient than Elements MC with average GI of 3.67(95% CI: 3.52–3.82) and 4.06 (95% CI: 3.93–4.20) respectively. Eclipse plans were more homogeneous with mean HI = 1.22 (95% CI: 1.20–1.23) that is lower than others. Average maximum clinical target volume (CTV) doses were higher in Elements MC with 22.31 Gy (95% CI: 21.87–22.74), while PB plans have 21.15 Gy (95% CI: 20.36–21.96), respectively. Elements MC and PB plans had lower average dose to 0.35 c.c. of spinal cord (D0.35cc) of 7.60 Gy (95% CI: 7.18–8.02) and 8.42 Gy (95% CI: 7.83–9.01). All plans had &gt;95% points passing the gamma QA criteria at 3%/2 mm. <jats:italic>Conclusion</jats:italic>. All treatment plans achieved clinically acceptable target coverage &gt;95% and meet spinal cord dose limits. Smart optimization in Brainlab Elements spine module produced dosimetrically superior plans by better spinal cord sparing.</jats:p>

<jats:title>Abstract</jats:title> <jats:p>The mechanism of the FLASH effect remains unclear and could be revealed by studying chemical reactions during irradiation. Monte Carlo simulation of the radiolytic species is an effective tool to analyze chemical reactions, but the simulation is limited by computing costs of the step-by-step simulation of radiolytic species, especially when considering beam with complex time structure. The complexity of the time structure of beams from accelerators in FLASH radiotherapy requires a high-performance Monte Carlo code. In this work, we develop a CPU-GPU coupling accelerating code with the independent reaction times (IRT) method to extend the chemical module of our nanodosimetry Monte Carlo code NASIC. Every chemical molecule in the microenvironment contains time information to consider the reactions from different tracks and simulate beams with complex time structures. Performance test shows that our code significantly improved the computing efficiency of the chemical module by four orders of magnitude. Then the code is used to study the oxygen depletion hypothesis in FLASH radiotherapy for different conditions by setting different parameters. The transient oxygen consumption rate values in the water are calculated when the pulses width ranges from 2 ps to 2 <jats:italic>μ</jats:italic>s, the total dose ranges from 0.5 Gy to 100 Gy and the initial oxygen concentration ranges from 0.1% to 21%. The time evolution curves are simulated to study the effect of the time structure of an electron linear accelerator. Results show that the total dose in several microseconds is a better indicator reflecting the radiolytic oxygen consumption rate than the dose rate. The initial oxygen greatly affects the oxygen consumption rate because of the reaction competition. The diffusion of oxygen determined by the physiological parameters is the key factor affecting oxygen depletion during the radiation using electron linear accelerators. Our code provides an efficient tool for simulating water radiolysis in different conditions.</jats:p>

<jats:title>Abstract</jats:title> <jats:p> <jats:italic>Purpose</jats:italic>. The Dynamic Collimation System (DCS) is an energy layer-specific collimation device designed to reduce the lateral penumbra in pencil beam scanning proton therapy. The DCS consists of two pairs of nickel trimmers that rapidly and independently move and rotate to intercept the scanning proton beam and an integrated range shifter to treat targets less than 4 cm deep. This work examines the validity of a single aperture approximation to model the DCS, a commonly used approximation in commercial treatment planning systems, as well as higher-order aperture-based approximations for modeling DCS-collimated dose distributions. <jats:italic>Methods</jats:italic>. An experimentally validated TOPAS/Geant4-based Monte Carlo model of the DCS integrated with a beam model of the IBA pencil beam scanning dedicated nozzle was used to simulate DCS- and aperture-collimated 100 MeV beamlets and composite treatment plans. The DCS was represented by three different aperture approximations: a single aperture placed halfway between the upper and lower trimmer planes, two apertures located at the upper and lower trimmer planes, and four apertures, located at both the upstream and downstream faces of each pair of trimmers. Line profiles and three-dimensional regions of interest were used to evaluate the validity and limitations of the aperture approximations investigated. <jats:italic>Results</jats:italic>. For pencil beams without a range shifter, minimal differences were observed between the DCS and single aperture approximation. For range shifted beamlets, the single aperture approximation yielded wider penumbra widths (up to 18%) in the X-direction and sharper widths (up to 9.4%) in the Y-direction. For the example treatment plan, the root-mean-square errors (RMSEs) in an overall three-dimensional region of interest were 1.7%, 1.3%, and 1.7% for the single aperture, two aperture, and four aperture models, respectively. If the region of interest only encompasses the lateral edges outside of the target, the resulting RMSEs were 1.7%, 1.1%, and 0.5% single aperture, two aperture, and four aperture models, respectively. <jats:italic>Conclusions</jats:italic>. Monte Carlo simulations of the DCS demonstrated that a single aperture approximation is sufficient for modeling pristine fields at the Bragg depth while range shifted fields require a higher-order aperture approximation. For the treatment plan considered, the double aperture model performed the best overall, however, the four-aperture model most accurately modeled the lateral field edges at the expense of increased dose differences proximal to and within the target.</jats:p>

<jats:title>Abstract</jats:title> <jats:p>Calcific aortic valve disease (CAVD) is the most common heart valvular disease in the developed world. Most of the relevant research has been sex-blind, ignoring sex-related biological variables and thus under-appreciate sex differences. However, females present pronounced fibrosis for the same aortic stenosis (AS) severity compared with males, who exhibit more calcification. Herein, we present a computational model of fibrocalcific AV, aiming to investigate its effect on AS development. A parametric study was conducted to explore the influence of the total collagen fiber volume and its architecture on the aortic valve area (AVA). Towards that goal, computational models were generated for three females with stenotic AVs and different volumes of calcium. We have tested the influence of fibrosis on various parameters as fiber architecture, fibrosis location, and transvalvular pressure. We found that increased fiber volume with a low calcium volume could actively contribute to AS and reduce the AVA similarly to high calcium volume. Thus, the computed AVAs for our fibrocalcific models were 0.94 and 0.84 cm<jats:sup>2</jats:sup> and the clinical (Echo) AVAs were 0.82 and 0.8 cm<jats:sup>2</jats:sup>. For the heavily calcified model, the computed AVA was 0.8 cm<jats:sup>2</jats:sup> and the clinical AVA was 0.73 cm<jats:sup>2</jats:sup>. The proposed models demonstrated how collagen thickening influence the fibrocalcific-AS process in female patients. These models can assist in the clinical decision-making process and treatment development in valve therapy for female patients.</jats:p>

<jats:title>Abstract</jats:title> <jats:p>Over the past few years, positron emission tomography/computed tomography (PET/CT) imaging for computer-aided diagnosis has received increasing attention. Supervised deep learning architectures are usually employed for the detection of abnormalities, with anatomical localization, especially in the case of CT scans. However, the main limitations of the supervised learning paradigm include (i) large amounts of data required for model training, and (ii) the assumption of fixed network weights upon training completion, implying that the performance of the model cannot be further improved after training. In order to overcome these limitations, we apply a few-shot learning (FSL) scheme. Contrary to traditional deep learning practices, in FSL the model is provided with less data during training. The model then utilizes end-user feedback after training to constantly improve its performance. We integrate FSL in a U-Net architecture for lung cancer lesion segmentation on PET/CT scans, allowing for dynamic model weight fine-tuning and resulting in an online supervised learning scheme. Constant online readjustments of the model weights according to the users’ feedback, increase the detection and classification accuracy, especially in cases where low detection performance is encountered. Our proposed method is validated on the Lung-PET-CT-DX TCIA database. PET/CT scans from 87 patients were included in the dataset and were acquired 60 minutes after intravenous <jats:sup>18</jats:sup>F-FDG injection. Experimental results indicate the superiority of our approach compared to other state-of-the-art methods.</jats:p>

<jats:title>Abstract</jats:title> <jats:p>Peripheral magnetic stimulation is a promising technique for several applications like rehabilitation or diagnose of neuronal pathways. However, most available magnetic stimulation devices are designed for transcranial stimulation and require high-power, expensive hardware. Modern technology such as rectangular pulses allows to adapt parameters like pulse shape and duration in order to reduce the required energy. Nevertheless, the effect of different temporal electromagnetic field shapes on neuronal structures is not yet fully understood. We created a simulation environment to find out how peripheral nerves are affected by induced magnetic fields and what pulse shapes have the lowest energy requirements. Using the electric field distribution of a <jats:italic>figure-of-8</jats:italic> coil together with an axon model in saline solution, we calculated the potential along the axon and determined the required threshold current to elicit an action potential. Further, for the purpose of selective stimulation, we investigated different axon diameters. Our results show that rectangular pulses have the lowest thresholds at a pulse duration of 20 <jats:italic>μ</jats:italic>s. For sinusoidal coil currents, the optimal pulse duration was found to be 40 <jats:italic>μ</jats:italic>s. Most importantly, with an asymmetric rectangular pulse, the coil current could be reduced from 2.3 kA (cosine shaped pulse) to 600 A. In summary, our results indicate that for magnetic nerve stimulation the use of rectangular pulse shapes holds the potential to reduce the required coil current by a factor of 4, which would be a massive improvement.</jats:p>

<jats:title>Abstract</jats:title> <jats:p>The problem of data truncation in Computed Tomography (CT) is caused by the missing data when the patient exceeds the Scan Field of View (SFOV) of a CT scanner. The reconstruction of a truncated scan produces severe truncation artifacts both inside and outside the SFOV. We have employed a deep learning-based approach to extend the field of view and suppress truncation artifacts. Thereby, our aim is to generate a good estimate of the real patient data and not to provide a perfect and diagnostic image even in regions beyond the SFOV of the CT scanner. This estimate could then be used as an input to higher order reconstruction algorithms [1]. To evaluate the influence of the network structure and layout on the results, three convolutional neural networks (CNNs), in particular a general CNN called ConvNet, an autoencoder, and the U-Net architecture have been investigated in this paper. Additionally, the impact of L1, L2, structural dissimilarity and perceptual loss functions on the neural network’s learning have been assessed and evaluated. The evaluation of data set comprising 12 truncated test patients demonstrated that the U-Net in combination with the structural dissimilarity loss showed the best performance in terms of image restoration in regions beyond the SFOV of the CT scanner. Moreover, this network produced the best mean absolute error, L1, L2, and structural dissimilarity evaluation measures on the test set compared to other applied networks. Therefore, it is possible to achieve truncation artifact removal using deep learning techniques.</jats:p>

<jats:title>Abstract</jats:title> <jats:p> <jats:italic>Purpose</jats:italic>: In radiotherapy, accuracy in dose estimation of dose calculation methods is critical. The influence of deformity on radiation dose calculations derived by planning system is evaluated in present study. The goal of study was to create a low-cost inhomogeneous phantom for measuring absorbed dose using an Ionisation chamber and Gafchromic film, which was validated using treatment planning system (TPS) dose outcome. <jats:italic>Methods:and Materials</jats:italic>: The central axis dose calculations were computed using Pencil Beam Convolution algorithm (PBC), Collapsed Cone Convolution (CCC) and Monte Carlo (MC) algorithm in the Monaco treatment planning system using an In-house phantom (20 × 20 × 20cm<jats:sup>3</jats:sup>) made up of acrylic sheet containing water and inhomogeneous material wooden powder equivalent to lung. Phantom was scanned in Computed Tomography (CT) scanner and image set was sent to the planning workstation. The depth dose evaluations were performed using ionization chamber and Gafchromic film with same beam settings and monitor units in every setup. Following that, the calculated doses obtained from TPS and measured depth doses were compared. <jats:italic>Results</jats:italic>: The results was reported for photon energies 6MV, 10MV, 15MV, 6FFF and 10FFF at varying field sizes of 4 × 4 cm<jats:sup>2</jats:sup>, 5 × 5 cm<jats:sup>2</jats:sup>, 10 × 10 cm<jats:sup>2</jats:sup>, and 15 × 15 cm<jats:sup>2</jats:sup>. MC maximum dose variation predicted was 2.06% in 15MV of measured chamber dose and −2.06% of measured gafchromic film dose in 6MVFFF. CCC maximum dose variation predicted was 2.68% of measured chamber dose in 6MV and 3.31% of measured gafchromic film dose in 6MV whereas PB maximum dose variation predicted was −5.94% in 15MV of measured chamber dose and −11.6% of measured gafchromic film dose in 6MVFFF. <jats:italic>Conclusion</jats:italic>: Low-cost in-house phantoms can be utilised to assess point and planar doses during patient-specific quality assurance in centres that don’t have accessibility to phantoms due to the high cost of commercially available tools.</jats:p>

<jats:title>Abstract</jats:title> <jats:p>In this study, the performance of a new iterative reconstruction algorithm, the pre-clinical AcurosXB iCBCT algorithm, has been characterized on Varian Halcyon linear accelerators with respect to the potential of radiotherapy dose calculations on CBCT images. The study utilized various phantom setups to verify the accuracy of the pre-clinical algorithm under different scatter conditions and compared dose calculations performed on CBCT images reconstructed with the pre-clinical algorithm to those performed on typical planning CT images. The results indicated that despite showing improvements compared to the existing iCBCT protocol, certain restrictions should be introduced when the pre-clinical AcurosXB iCBCT algorithm was used for dose calculations. Changes in the scatter condition exhibited a larger effect on CBCTs than on planning CTs. Therefore, users should be careful in offsetting the patient and positioning the patient’s arms if the resultant images will be used for dose calculations. In addition, protocols with different kV settings should be approached with caution, where 100 kV protocols should only be used to scan the head and neck area, while the rest of the body should be scanned with the 125 kV and 140 kV protocols. When the patient is set up properly and the appropriate energy is selected for the anatomical area, the uncertainty of using the novel AcurosXB iCBCT algorithm for treatment planning dose calculation is within ±2.0%.</jats:p>