Catálogo de publicaciones - revistas

<jats:title>Abstract</jats:title> <jats:p>Objective: The goal of this study was to use Monte Carlo (MC) simulations and measurements to investigate the dosimetric suitability of an interventional radiology (IR) c-arm fluoroscope to deliver low-dose radiotherapy to the lungs. Approach: A previously-validated MC model of an IR fluoroscope was used to calculate the dose distributions in a COVID-19-infected patient, 20 non-infected patients of varying sizes, and a postmortem subject. Dose distributions for PA, AP/PA, 3-field and 4-field treatments irradiating 95% of the lungs to a 0.5 Gy dose were calculated. An algorithm was created to calculate skin entrance dose as a function of patient thickness for treatment planning purposes. Treatments were experimentally validated in a postmortem subject by using implanted dosimeters to capture organ doses. Main Results: Mean doses to the left/right lungs for the COVID-19 CT data were 1.2/1.3 Gy, 0.8/0.9 Gy, 0.8/0.8 Gy and 0.6/0.6 Gy for the PA, AP/PA, 3-field, and 4-field configurations, respectively. Skin dose toxicity was the highest probability for the PA and lowest for the 4-field configuration. Dose to the heart slightly exceeded the ICRP tolerance; all other organ doses were below published tolerances. The AP/PA configuration provided the best fit for entrance skin dose as a function of patient thickness (R2 = 0.8). The average dose difference between simulation and measurement in the postmortem subject was 0.7%. Significance: An IR fluoroscope should be capable of delivering low-dose radiotherapy to the lungs with tolerable collateral dose to nearby organs.</jats:p>

<jats:title>Abstract</jats:title> <jats:p>Background With the increasing survival of the congenital heart disease population, there is a growing need for in-depth understanding of blood circulation in these patients. Mock loops provide the opportunity for comprehensive hemodynamic studies without burden and risks for patients. This study aimed to evaluate the ability of the presented mock loop to mimic the hemodynamics of the pulmonary circulation with and without stenosis and the MR compatibility of the system. Methods A pulsatile pump with two chambers, separated by a flexible membrane, was designed and 3D printed. A cough assist device applied an alternating positive and negative pressure on the membrane. One adult, and three pediatric pulmonary bifurcations were 3D printed and incorporated in the setup. Two pediatric models had a 50% stenosis of the left branch. Bilateral compliance chambers allowed for individual compliance tuning. A reservoir determined the diastolic pressure. Two carbon heart valves guaranteed unidirectional flow. The positive pressure on the cough assist device was tuned until an adequate stroke volume was reached with a frequency of 60 bpm. Flow and pressure measurements were performed on the main pulmonary artery and the two branches. The MR compatibility of the setup was evaluated. Results A stroke volume with a cardiac index of 2L/min/m2 was achieved in all models. Physiological pressure curves were generated in both normal and stenotic models. The mock loop was MR compatible. Conclusion This MR compatible mock loop, closely resembles the pulmonary circulation thereby providing a controllable environment for hemodynamic studies. </jats:p>

<jats:title>Abstract</jats:title> <jats:p>Objective: In this modelling study, we pursued two main goals. The first was to establish a new CTV-to-PTV expansion which considers the closest and most critical organ at risk (OAR). The second goal was to investigate the impact of the planning target volume (PTV) margin size on the tumor control probability (TCP) and its dependence on the geometrical setup uncertainties. The aim was to achieve a smaller margin expansion close to the OAR while allowing a moderately larger expansion in less critical areas further away from the OAR and whilst maintaining the TCP. Approach: Imaging data of radiation therapy plans from pet dogs which had undergone radiation therapy for brain tumor were used to estimate the clinic specific rotational setup uncertainties. A Monte-Carlo methodology using a voxel-based TCP model was used to quantify the implications of rotational setup uncertainties on the TCP. A combination of algorithms was utilized to establish a computational CTV-to-PTV expansion method based on probability density. This was achieved by choosing a center of rotation close to an OAR. All required software modules were developed and integrated into a software package that directly interacts with the Varian Eclipse treatment planning system. Main results: Several uniform and non-isotropic PTVs were created. To ensure comparability and consistency, standardized RT plans with equal optimization constraints were defined, automatically applied and calculated on these targets. The resulting TCPs were then computed, evaluated and compared. Significance: The non-isotropic margins were found to result in larger TCPs with smaller margin excess volume. Further, we presented an additional application of the newly established CTV-to-PTV expansion method for radiation therapy of the spinal axis of human patients.</jats:p>

<jats:title>Abstract</jats:title> <jats:p>Bacterial infections have a large impact on public health. Through this study, we report on the development of complementary split-ring resonators (CSRR) supplemented by functionalized nanoparticles to detect bacteria in the aqueous medium. Iron oxide (Fe<jats:sub>3</jats:sub>O<jats:sub>4</jats:sub>) nanoparticles were functionalized with amino groups using (3-aminopropyl) triethoxysilane (APTES) to form (APTES@Fe<jats:sub>3</jats:sub>O<jats:sub>4</jats:sub>) nanoparticles, which have a specific affinity towards the bacterial species. This affinity was evaluated using the <jats:italic>Escherichia coli (E. coli)</jats:italic> and <jats:italic>Staphylococcus aureus (S. aureus)</jats:italic> bacterial species. The resonant sensor was tuned at 430 MHz and the CSRR sensor bed was further activated using APTES@Fe<jats:sub>3</jats:sub>O<jats:sub>4</jats:sub> nanoparticles. Bacterial detection was studied over a range of concentrations from 2.66 x 10<jats:sup>9</jats:sup> cells to 2.66 x 10<jats:sup>8</jats:sup> cells. The sensor actively responded to small changes in bacterial concentration, showing an overall shift in resonance frequency of ~ 44 MHz (~ 40 MHz / cell count) for <jats:italic>E. coli</jats:italic> and ~ 55 MHz (50.43 MHz / cell count) for <jats:italic>S. aureus</jats:italic>. Dextran sulphate and Chitosan were used as the references. The magnetic character of the conjugated system exhibited strong interaction of the bacterial species with APTES@Fe<jats:sub>3</jats:sub>O<jats:sub>4</jats:sub>, justifying the high selectivity towards these species. This demonstrates the feasibility of a sensitive, fast, portable device, against the traditionally used time-consuming bio-assays.</jats:p>

<jats:title>Abstract</jats:title> <jats:p>This paper presents the synthesis of a mixture solution that is equivalent to ex-vivo liver tissue dielectric characteristics between 500 MHz and 5 GHz. The mimicking solution was synthesized using concentrations of two chemicals, the solute which is referred to as the inclusion phase and the solvent, referred to as the host phase. The inclusion phase consisted of bovine serum albumin (BSA) powder and the host phase was comprised of a phosphate-buffered saline (PBS) solution with a concentration of Triton X-100 (TX-100). The dielectric properties of these two phases were substituted into Bruggeman’s two-phase mixture equation to estimate the dielectric properties of an excised liver tissue. Furthermore, the study exploits Bruggeman’s equation to investigate the impact of tissue dehydration levels on the dielectric properties of an excised tissue. The effect of dehydration has been characterised as a function of time based on the loss-on-drying technique (a substance is heated until it is completely dry). Dielectric parameters were measured as a function of frequency using a slim form open-ended coaxial probe at a constant room temperature of circa 25◦ C. Measured dielectic data were fitted to the Cole-Cole model and good agreement with the mimicking solutions was obtained. These results indicate that these solutions can be used to model the human body phantoms for microwave medical applications.</jats:p>

<jats:title>Abstract</jats:title> <jats:p>Power efficiency in electrical stimulator circuits is crucial for developing large-scale multichannel applications like bidirectional brain-computer interfaces and neuroprosthetic devices. Many state-of-the-art papers have suggested that some non-rectangular pulse shapes are more energy-efficient for exciting neural excitation than the conventional rectangular shape. However, additional losses in the stimulator circuit, which arise from employing such pulses, were not considered. In this work, we analyze the total energy efficiency of a stimulation system featuring non-rectangular stimuli, taking into account the losses in the stimulator circuit. To this end, activation current thresholds for different pulse shapes and durations in cortical neurons are modeled, and the energy required to generate the pulses from a constant voltage supply is calculated. The proposed calculation reveals an energy increase of 14-51% for non-rectangular pulses compared to the conventional rectangular stimuli, instead of the decrease claimed in previous literature. This result indicates that a rectangular stimulation pulse is more power-efficient than the tested alternative shapes in large-scale multichannel electrical stimulation systems.</jats:p>

<jats:title>Abstract</jats:title> <jats:p>This paper proposes the transition times of Petri net models of human gait as training features for multiclass random forests (RFs) and classification trees (CTs). These models are designed to support screening for neurodegenerative diseases. The proposed Petri net describes gait in terms of nine cyclic phases and the timing of the nine events that mark the transition between phases. Since the transition times between strides vary, each is represented as a random variable characterized by its mean and standard deviation. These transition times are calculated using the PhysioNet database of vertical ground reaction forces (VGRFs) generated by feet-ground contact. This database comprises the VGRFs of four groups: amyotrophic lateral sclerosis, the control group, Huntington’s disease, and Parkinson disease. The RF produced an overall classification accuracy of 91%, and the specificities and sensitivities for each class were between 80% and 100%. However, despite this high performance, the RF-generated models demonstrated lack of interpretability prompted the training of a CT using identical features. The obtained tree comprised only four features and required a maximum of three comparisons. However, this simplification dramatically reduced the overall accuracy from 90.6% to 62.3%. The proposed set features were compared with those included in PhysioNet database of VGRFs. In terms of both the RF and CT, more accurate models were established using our features than those of the PhysioNet.</jats:p>

<jats:title>Abstract</jats:title> <jats:p>A phantom is a highly specialized device, which mimic human body, or a part of it. There are three categories of phantoms: physical phantoms, physiological phantoms, and computational phantoms. The phantoms have been utilized in medical imaging and radiotherapy for numerous applications. In radiotherapy, the phantoms may be used for various applications such as quality assurance (QA), dosimetry, end-to-end testing, etc In thoracic radiotherapy, unique QA problems including tumor motion, thorax deformation, and heterogeneities in the beam path have complicated the delivery of dose to both tumor and organ at risks (OARs). Also, respiratory motion is a major challenge in radiotherapy of thoracic malignancies, which can be resulted in the discrepancies between the planned and delivered doses to cancerous tissue. Hence, the overall treatment procedure needs to be verified. Anthropomorphic thorax phantoms, which are made of human tissue-mimicking materials, can be utilized to obtain the ground truth to validate these processes. Accordingly, research into new anthropomorphic thorax phantoms has accelerated. Therefore, the review is intended to summarize the current status of the commercially available and in-house-built anthropomorphic physical/physiological thorax phantoms in radiotherapy. The main focus is on anthropomorphic, deformable thorax motion phantoms. This review also discusses the applications of three-dimensional (3D) printing technology for the fabrication of thorax phantoms.</jats:p>

<jats:title>Abstract</jats:title> <jats:p> <jats:italic>Objective</jats:italic>. Auto-contouring of organs at risk (OAR) is becoming more common in radiotherapy. An important issue in clinical decision making is judging the quality of the auto-contours. While recent studies considered contour quality by looking at geometric errors only, this does not capture the dosimetric impact of the errors. In this work, we studied the relationship between geometrical errors, the local dose and the dosimetric impact of the geometrical errors. <jats:italic>Approach</jats:italic>. For 94 head and neck patients, unmodified atlas-based auto-contours and clinically used delineations of the parotid glands and brainstem were retrieved. VMAT plans were automatically optimized on the auto-contours and evaluated on both contours. We defined the dosimetric impact on evaluation (DIE) as the difference in the dosimetric parameter of interest between the two contours. We developed three linear regression models to predict the DIE using: (1) global geometric metrics, (2) global dosimetric metrics, (3) combined local geometric and dosimetric metrics. For model (3), we next determined the minimal amount of editing information required to produce a reliable prediction. Performance was assessed by the root mean squared error (RMSE) of the predicted DIE using 5-fold cross-validation. <jats:italic>Main results</jats:italic>. In model (3), the median RMSE of the left parotid was 0.4 Gy using 5% of the largest editing vectors. For the right parotid and brainstem the results were 0.5 Gy using 10% and 0.4 Gy using 1% respectively. The median RMS of the DIE was 0.6 Gy, 0.7 Gy and 0.9 Gy for the left parotid, the right parotid and the brainstem, respectively. Model (3), combining local dosimetric and geometric quantities, outperformed the models that used only geometric or dosimetric information. <jats:italic>Significance</jats:italic>. We showed that the largest local errors plus the local dose suffice to accurately predict the dosimetric impact, opening the door to automated dosimetric QA of auto-contours.</jats:p>

<jats:title>Abstract</jats:title> <jats:p>Previously, convolutional neural networks mostly used deep semantic feature information obtained from several convolutions for image classification. Such deep semantic features have a larger receptive field, and the features extracted are more effective as the number of convolutions increases, which helps in the classification of targets. However, this method tends to lose the shallow local features, such as the spatial connectivity and correlation of tumor region texture and edge contours in breast histopathology images, which leads to its recognition accuracy not being high enough. To address this problem, we propose a multi-level feature fusion method for breast histopathology image classification. First, we fuse shallow features and deep semantic features by attention mechanism and convolutions. Then, a new weighted cross entropy loss function is used to deal with the misjudgment of false negative and false positive. And finally, the correlation of spatial information is used to correct the misjudgment of some patches. We have conducted experiments on our own datasets and compared with the base network Inception-ResNet-v2, which has a high accuracy. The proposed method achieves an accuracy of 99.0% and an AUC of 99.9%.</jats:p>